Skip to main content

Full text of "Smokeless powder, nitro-cellulose, and theory of the cellulose molecule"

See other formats


,,,:,,, N ^ 


Acce&sio u No . 92 3 2 3 >> Class No . 






Lieutenant, United States Navy 






Copyright, 1901, 




FOR purposes of comparative study, the writer has 
brought together in the present volume a series of 
papers, by various investigators, upon the composition 
of cellulose and the properties of explosives prepared 
therefrom. He has supplemented these with an ac- 
count of experiments made by himself; and from the 
whole has drawn certain conclusions as to the possible 
ultimate chemical composition of cellulose and the 

While the general development of war-material 
from the mechanical and metallurgical standpoints 
the production of ordnance and armor is so largely 
identified with progress in the useful arts in the 
United States, yet, until very recently, but little has 
been accomplished in our country in the way of im- 
provement in explosives. Within the last few years, 
however, a particular form of smokeless powder has 
largely supplanted the old black and brown powders 
for military uses ; and the last decade of the past 
century has witnessed the virtual abandonment of 
a propellant that has held its place in war, with 




comparatively little modification, for four hundred 

This new smokeless powder, which is adapted for 
use in arms of all calibres, is prepared from a particu- 
lar type of colloid nitro-cellulose. Such an extension 
of the employment of this latter body from its original 
use for detonating purposes, to its new use as a pro- 
gressive explosive, has attracted general attention, and 
led to a more careful and extended study of the nitro- 
celluloses in general. It is with the view of further 
extending such study and of possibly preparing the 
way for the introduction of future improvements in 
progressive explosives that this book has been pre- 

Before presenting it, the author wishes to. express 
his thanks to certain eminent scientists for the privi- 
lege that they have courteously afforded him of 
making his own translations of certain portions of their 
works upon explosives : to Professor D. Mendeleef, 
of Russia, for his paper entitled " Pyrocollodion 
Smokeless Powders"; to M. Vieille, of the French 
Service des Poudres et Salpetres, for his article upon 
the nitration of cotton ; and to M. Bruley, of the same 
service, for a similar paper. Thanks are due also to 
Messrs. Longmans, Green & Co. for the privilege 
kindly extended of making certain extracts from 
Messrs. Cross and Bevan's valuable work, ' ' Cellulose, ' ' 
published by them. 

Finally, the author wishes to express his indebted- 
ness to Dr. Alfred I. Cohn, of New York, for the care 
he has bestowed upon the reading of the proof of 


this book at a time when the writer's absence from 
the United States on active service afloat prevented 
his giving the matter the personal care and attention 
it otherwise would have had ; and for his prepara- 
tion of a comprehensive index. 

U. S. S. " Dixie," April 24, 1901. 




Origin 1 

Nomenclature , 4 

Definitions 5 










I. First Method of Nitration 81 

II. Second Method of Nitration 87 

III. Higher Limit of Degree of Nitration 91 

IV. Cellulose Incompletely Nitrated 92 

V. Resum6 and Conclusions. . 94 







I. Method of NitraCion and Mode of Representing Results. . 130 

II. Results Obtained by M. Vieille 131 

III. Experiments in Nitration (First Series) 134 

IV. Experiments in Nitration (Second Series) 140 

V. Experiments in Nitration (Third Series) 141 

VI. Various Experiments 144 

VII. Resume and Conclusions 152 



INDEX .. . 195 



THE discovery that cellulose, by treatment with 
nitric acid, is converted into a highly inflammable or 
explosive body was made during the first half of the 
nineteenth century. The action of nitric acid on 
starch was investigated to some extent in 1833 by 
Braconnot, who found that a very rapidly burning 
material was produced, and which he named xyloidine. 
Pelouze further investigated this substance in 1838, 
and also studied similar bodies prepared from paper, 
linen, etc., which he held to be identical with the one 
from starch. 

In 1846 Schonbein discovered that cellulose in the 
form of cotton, when immersed in nitric acid, freed 
from the acid excess, and dried, was converted into a 
highly explosive compound. This latter substance, 
subsequently known as gun-cotton, constitutes the 
base of modern smokeless powders.* 

* The name " gun-cotton," as originally employed, was generic 
for all varieties of the more highly explosive nitro-celluloses pre- 
pared from cotton. 


On its first appearance gun-cotton was hailed by 
ordnance experts as a smokeless substitute for gun- 
powder, and extended series of experiments were con- 
ducted to prove its adaptability to the requirements of 
war. The combined efforts of chemist, powder-maker, 
and artillerist failed, however, to secure such a control 
of its combustion as would enable it to be safely em- 
ployed, in charges of a given weight, to develop a uni- 
form bore-pressure with standard projectiles in guns of 
a given calibre. After the occurrence of a number of 
detonations, unlocked for and more or less disastrous, 
its use was abandoned. Nevertheless, the study of 
the composition, properties, and methods of production 
of gun-cotton and kindred bodies continued to be sys- 
tematically prosecuted, the theory of nitro-substitution 
was evolved, and the knowledge of the applicability of 
the new materials to military requirements extended. 
But more than forty years were to elapse from the 
time of the first experiments before a thoroughly re- 
liable smokeless powder, free from all tendency to 
detonation, producing minimum erosion, of good 
keeping qualities, and of maximum propulsive effect, 
was to be obtained. 

To trace the various steps of progress in the deter- 
mination of the properties of gun-cotton is exceed- 
ingly difficult. It is primarily difficult for the reason 
that the base-material under investigation possesses 
an organic structure ; and because the resultant nitrated 
product varies in chemical composition according to 
strength and temperature of the acids employed in 
its preparation, their water content, duration of re- 


action, and temperature at which the reaction is con- 
ducted. It is difficult secondarily for the reason that 
the ultimate product of nitration is either resolvable, 
by treatment with certain solvents, into different sub- 
components, or else is capable of direct preparation in 
a form soluble in one or another solvent; and because 
certain industries, sciences, and professions the man- 
ufacture of explosives, of fabrics, photography, sur- 
gery have required the development of special forms 
of material and demanded special lines of research. 

The existence to-day of a very large number of 
unclassified names to denote the different varieties of 
nitro-cellulose, serves as an illustration of the necessa- 
rily involved and complex nature of such investigations 
as have been made for the purpose of determining its 
composition, and of the incompleteness of these inves- 
tigations. The confusion in nomenclature that has 
arisen is attributable to the fact that the formula for 
cellulose, and therefore for the cellulose nitrates, not 
being established, investigators in different countries 
started in independently to determine simultaneously 
both the properties of these compounds and their 
chemical constitutions. Matters became further in- 
volved through the efforts to translate the accounts of 
the work of one investigator into the language of 

With the view of avoiding confusion in the future it 
may be well to anticipate somewhat here, to indicate 
the origin of the various names applied to forms of 
nitro-cellulose, to define and classify them, and to 
indicate such of them as will be employed hereafter in 


the body of the present work ot denote those distinct 
varieties that possess significance from the standpoint 
of explosive effect. 


Names may be divided into classes, according to 
origin, as follows: 

i. Those devised by discoverers to denote new 
chemical compounds. Thus, xyloidin, from uAos", 
wood ; pyroxylin, from nvp, fire, and uAo?, wood. 

2. Those implying uses to which special forms of 
the material are applied, as gun-cotton, collodion-cot- 

3. - 1 Those originating in references to the theory 
of nitro--substitution ; thus, nitro-cellulose, mono-, di-, 
tri-, tetra-, penta-, etc., nitro-cellulose. 

4. Names based upon the physical characteristics 
of the material, as "soluble nitro-cellulose," "insol- 
uble nitro-cellulose," " friable cottons " (Vieille). 

5. Names obviously incorrect as to application and 
meaning, as " soluble cotton " for soluble -nitre-cellu- 
lose; " insoluble cotton " for insoluble nitro-cellulose. 

6. Generic names for groups of varieties between 
which differences in chemical composition were recog- 
nized as existing; thus, nitro-celluloses as opposed to 

7. Names referring to chemical composition as 
modified by extent and character of nitro-substitution, 
as "soluble nitro-cellulose of high nitration," of "low 



Cellulose. The cell-wall or envelope of plant-tissues, 
to which the name cellulose has been applied as to a 
chemical individual. Unless the contrary is stated, 
the term cellulose, as employed in the present work, 
refers to pure cotton, unbleached and unspun, freed by 
mechanical treatment (ginning, picking, boiling, etc.) 
from wood, dirt, greases, resins, and foreign matter in 
general; either in the natural state, or as waste product 
from industrial processes. 

Nitration. The displacement of a number of atoms 
of replaceable hydrogen in cellulose, and the substitu- 
tion therefor of the univalent radicle nitryl (NO a ). 
The term " nitration " is also employed to indicate the 
percentage of nitrogen in a given nitro-cellulose. 

Nitro-cellulose. Products resulting from the treat- 
ment of cellulose with strong nitric acid under the 
condition that they retain the cellular structure of the 
original cotton. 

Nitro-cellulose of high nitration. Those forms of 
nitro-cellulose in which a relatively large number of 
the replaceable hydrogen atoms are replaced by nitryl. 

Nitro-cellulose of mean nitration. Those forms of 
nitro-cellulose in which a mean number of the replace- 
able hydrogen atoms are replaced by nitryl. 

Nitro-cellulose of low nitration. Those forms of 
nitro-cellulose in which a relatively small number of 
replaceable hydrogen atoms are replaced by nitryl. 

Insoluble nitro-cellulose. Those forms of nitro-cel- 
lulose of high nitration insoluble at ordinary atmos- 


pheric temperatures in a mixture of two parts by 
weight of ethyl ether and one part by weight of ethyl 

Soluble nitro-celliilose. Those forms of nitro-cellu- 
lose of low or mean nitration soluble at ordinary at- 
mospheric temperature in a mixture of two parts by 
weight of ethyl ether and one part by weight of ethyl 

Hydrocellulose. The product obtained by exposing 
cotton to the action of hydrochloric-acid fumes ; or by 
immersing it in hydrochloric, dilute sulphuric, or very 
dilute nitric acid ; a white pulverulent mass which, 
under the microscope, is seen to consist of fragments of 
the original fibre of modified cellular form. 

Nitro-hydrocellulose. Products resulting from the 
treatment of hydrocellulose with strong nitric acid 
under the condition that the resultant product retains 
the cellular structure originally possessed by the hy- 

Nitro-hydrocellulose of high nitration. Those forms 
of nitro-hydrocellulose in which a relatively large 
number of the replaceable hydrogen atoms are replaced 
by nitryl. 

Nitro-hydrocellulose of mean nitration. Those 
forms of nitro-hydrocellulose in which a mean number 
of the replaceable hydrogen atoms are replaced by 

Nitro-hydrocellulose of low nitration. Those forms 
of nitro-hydrocellulose in which a relatively small 
number of replaceable hydrogen atoms are replaced by 


Insoluble nitro-Jiydroccllulose. Those forms of nitro- 
hydrocellulose of high nitration insoluble at ordinary 
atmospheric temperatures in a mixture of two parts 
by weight of ethyl ether and one part by weight of 
ethyl alcohol. 

Soluble nitro-hydrocellulose. Those forms of nitro- 
hydrocellulose of low or mean nitration soluble at 
ordinary atmospheric temperatures in a mixture of 
two parts by weight of ethyl ether and one part by 
weight of ethyl alcohol. 

Gun-cotton. The military name for those forms of 
highly explosive nitro-celluloses employed in war, and 
which are generally mixtures of a large quantity of in- 
soluble with a small quantity of soluble nitro-cellulose 
and a very small quantity of unnitrated cotton. 

Pyrocellulose. Soluble nitro-cellulose of high uni- 
form nitration, possessing a sufficient content of oxy- 
gen to convert its carbon into carbonic oxide and its 
hydrogen into aqueous vapor. 

These terms, as defined above, will be employed so 
far as possible in the body of this treatise. The follow- 
ing list of synonyms is supplied for reference : 


Name Synonym 

Nitro-cellulose, Pyroxyline or pyroxylin. 

Insoluble nitro-cellulose, Insoluble gun-cotton ; insolu- 

ble cotton. 

Soluble nitro-cellulose, Soluble gun-cotton; soluble 

cotton; collodion-cotton; col- 

Soluble nitro-cellulose of low Friable cotton. 



EXISTING knowledge of the composition and consti- 
tution of nitro-cellulose still remains in a state of great 
confusion. A number of papers, scattered through- 
out the literature of experimental chemistry, and 
which throw valuable light upon the subject, have been 
published by independent investigators, while from 
time to time more elaborate articles summarizing the 
results of early workers have appeared. At least one 
excellent treatise upon cellulose has been published, * 
but it devotes only a few pages to the consideration 
of the nitro-celluloses. The reason for the existing 
confusion is that as yet no definite chemical structure 
has been determined for nitro-celluloses; they are re- 
garded as nitro-substitution compounds, as ethers, or 
else their composition is considered as still doubtful. 

If, however, the various original experimental re- 
sults and the conclusions that have been drawn there- 
from be examined in sequence, it will be observed that 
there does exist a tendency to account for the compo- 
sition of these bodies on a definite hypothesis. There- 

*" Cellulose," Cross and Bevan. London: Longmans, Green 
& Co., 1895. 



fore the writer, in his endeavor to throw some light 
upon ultimate cellulose- and nitro-cellulose composi- 
tion, will take up for consideration, first of all, in 
natural order of succession, results obtained by a 
number of students, each of whom, in his work, has 
supplied material from which subsequent investigators 
have drawn important conclusions. 

Almost from the beginning of the study of the body 
the existence of more than one form was recognized. 
Domonte and Menard's discovery that pyroxylin of 
low nitration was soluble in ether-alcohol, while the 
more highly nitrated variety remained insoluble there- 
in, furnished a positive differentiation of nitro-cellu- 
loses into two groups; Bechamp showed that there 
existed nitro-celluloses of different nitrations soluble 
in ether-alcohol ; and the only way of explaining the 
existence of such differences, in accordance with 
chemical theory, was by assuming that the substances 
of varying degrees of nitration obtained consisted of 
mixtures of various quantities of different, definite, 
chemical compounds. 

The numerous attempts to explain both the mode 
of formation of nitro-cellulose as well as its constitu- 
tion, and to reconcile results from its analyses, are well 
illustrated by the formulae selected by early investi- 
gators to represent its varieties. Of these formulas, 
numerous series exist, extending from the original 
formula of Schonbein, through the later formulated, 
so-called mono-, di-, and trinitro-celluloses, to the 
series of six nitrates of Eder. At an early date the 
belief became established, based in all probability upon 


analogies furnished by the nitre-glycerins and nitro- 
benzols, that nitro-celluloses were mixtures of nitro- 
substitution products, which, assuming the composition 
of cellulose as C 6 H 10 O B , were formulated as the tri-, di-, 
and mononitrates, with compositions C 6 H,O 8 (NO a ),, 
C 8 H 8 O 6 (NO a ) 3 , and C 6 H 9 O 6 (NO,), respectively. Of 
these, trinitro-cellulose was considered as identical 
with the insoluble variety of high nitration, and di- 
nitro-cellulose with that soluble in ether-alcohol; the 
existence of the mononitro-cellulose was predicated. 
Certain reactions due to these bodies led to a change 
of views concerning their chemical structure. B6champ 
found that nitro-celluloses dissolved in ether-alcohol 
surrendered nitric acid upon addition of potash or 
ammonia, with the resultant formation of nitro-cellu- 
loses of lower nitration ; he claimed for them the 
composition of ethers.* 

Other reactions of the nitro-celluloses tend to sup- 
plement this view of their composition; e.g., ferric 
chloride and potassium and ammonium sulphydrates 

* An ether is one of a class of organic bodies divided into two 
groups : (i) simple ethers, consisting of two basic hydrocarbon 
radicles united by oxygen, and corresponding in constitution to 
the metallic oxides, as CH 3 OCH 3 , methyl ether, or methyl oxide, 
analogous to AgOAg, silver oxide ; (2) compound ethers, consist- 
ing of one or more basic or alcohol radicles and one or more acid 
or hydrocarbon radicles united by oxygen, and corresponding to 
the salts of the metals, as CH 3 COOC 3 H 5 , ethyl acetate or acetic 
ether, corresponding to CHsCOONa, sodium acetate. Thus, if a 
cellulose be represented as possessing a composition C 6 Hi O 6 , 
and be written as a tribasic alcohol, CH 7 O 2 (OH) 3 , then, on sub- 
stituting nitryl, Nd, for the replaceable hydrogen, we obtain 
, a compound nitric ether. 


(Hadow, von Pettenkofer, cited by Guttmann) occasion 
the recovery of their cellulose, while the liberated 
nitric acid oxidizes the iron of the ferric chloride and 
transforms the sulphydrates into nitrates. 

In 1878 Dr. J. M. Eder conducted a series of in- 
vestigations into the character and composition of the 
nitro celluloses which has had much influence upon 
subsequent formation of thought in relation to the 
character and composition of these bodies; and from 
the results of his labors he was led to conclude that 
there existed no less than six distinct varieties of them, 
three of which, the hexa-, penta-, and di-, he was able 
to isolate ; two of them, the tetra- and tri-, he obtained 
in admixture; the mononitro-cellulose, however, he 
was unable to prepare. Doubling the coefficients of 
cellulose to avoid fractional coefficients in the deriva- 
tives, he formulated his series of nitrates as follows : 

Name Composition 

Cellulose hexanitrate CiaHi4O 4 (NO 3 ) 

Cellulose pentanitrate C, 3 H 16 O 5 (NO3)6 

Cellulose tetranitrate CiaHieC^NOa^ 

Cellulose trinitrate Ci 3 Hi7O7(NO)s 

Cellulose dinitrate CnH 16 O 8 (NO 3 ) a 

Cellulose mononitrate 

Dr. Eder's work constituted a purely scientific re- 
search into the chemical properties of a group of 
bodies. Subsequent investigations are characterized 
by the partial subordination of their scientific aims to 
the demands of the useful arts. These investigations 
are of two kinds, and refer to the technical uses of the 
soluble and insoluble varieties respectively. Those 


upon the soluble nitro-celluloses relate, for example, to 
photography, to questions of transparency and uni- 
formity of thickness of film ; those upon insoluble nitro- 
celluloses, to the art of war and to the attainment of 
the highest explosive effect consistent with the main- 
tenance of stability. The general purpose of the 
present work is the study of the physical and chemical 
properties of nitro-celluloses and nitro-cellulose col- 
loids, their methods of preparation, and their explosive 
qualities; and to the prosecution of this study investi- 
gations of the second class have, until recently, 
afforded the more direct aid. Conditions of research 
have, however, been recently modified. For, whereas 
the explosives prepared from cotton were formerly 
insoluble nitro-celluloses (gun-cottons) susceptible of 
detonation as well as of combustion, and employed in 
mines and torpedoes, yet in recent times the field of 
research has been extended to the soluble varieties in 
the effort to secure a suitable base material for the 
preparation of anon-detonating, progressively-burning, 
smokeless powder. 

The following account of Eder's nitrates (taken 
from Cross and Bevan, " Cellulose " *) may be quoted 
here : 

" Hexanitrate, C 12 H 14 O 4 (NO 3 ) 6 , gun-cotton. In 
the formation of this body, nitric acid of 1.5 sp. gr. 
and sulphuric acid of 1.84 sp. gr. are mixed in vary- 
ing proportions, about 3 of nitric to I of sulphuric ; 
sometimes this proportion is reversed, and cotton 

* By permission of Longmans, Green & Co. 


immersed in this at a temperature not exceeding 
10 C. for 24 hours; 100 parts of cellulose yield 
about 175 of cellulose nitrate. The hexanitrate so 
prepared is insoluble in alcohol, ether, or mixtures 
of both, in glacial acetic acid or methyl alcohol. 
Acetone dissolves it very slowly. This is the most 
explosive gun-cotton. It ignites at i6o-i7o C. 
According to Eder the mixtures of nitre and sul- 
phuric acid do not give this nitrate. Ordinary gun- 
cotton may contain as much as 12 per cent, of nitrates 
soluble in ether-alcohol. The hexanitrate seems to 
be the only one quite insoluble in ether-alcohol." 

" Pentanitrate, C 12 H 16 O 6 (NO 8 ) 6 . This composition 
has been very commonly ascribed to gun-cotton. 
It is difficult, if not impossible, to prepare it in a state 
of purity by the direct action of acid on cellulose. 
The best method is the one devised by Eder, making 
use of the property discovered by De Vrij, that gun- 
cotton (hexanitrate) dissolves in nitric acid at about 
80 or 90 C., and is precipitated, as the penta- 
nitrate, by concentrated sulphuric acid after cooling 
to o C. ; after mixing with a large volume of water, 
and washing the precipitate with water and then with 
alcohol, it is dissolved in ether-alcohol and again pre- 
cipitated with water, when it is obtained pure. This 
nitrate is insoluble in alcohol, but dissolves readily 
in ether-alcohol, and slightly in acetic acid. Strong 
potassa solution converts this nitrate into the di- 
nitrate, C 1 .H,.O i (NO.) 1 ." 

"The tetra- and trinitrates (collodion-pyroxylin) 
are generally formed together when cellulose is 


treated with a more dilute nitric acid, and at a higher 
temperature, and for a much shorter time (13-20 
minutes), than in the formation of the hexanitrate. 
It is not possible to separate them, as they are sol- 
uble to the same extent in ether-alcohol, acetic ether, 
acetic acid, or wood-spirit. On treatment with con- 
centrated nitric and sulphuric acids, both the tri- and 
tetranitrates are converted into pentanitrate and hexa- 
nitrate. Potassa and ammonia convert them into 

"Cellulose dinitrate, C ia H 18 O 8 (NO,) a , is formed 
by the action of alkalies on the other nitrates, and 
also by the action of hot dilute nitric acid on cellu- 
lose. The dinitrate is very soluble in alcohol-ether, 
acetic ether, and in absolute alcohol. Further action 
of alkalies on the dinitrate results in a complete 
decomposition of the molecule, some organic acids 
and tarry matters being formed." 

The next after Eder to increase our knowledge of 
the character and relationships of the members of the 
nitro-cellulose series was M. Vieille (Comptes Rendus, 
95, 132), the eminent French savant, whose efforts (in 
collaboration with M. Sarrau) were crowned with 
success in the production and establishment of 
the manufacture of a successful smokeless powder in 
France. M. Vieille's researches were published in 
part in the French official journal of explosives, the 
Memorial des Poudres et Salpetres,* a translation of 

* " Recherches sur la Nitrification du Coton," by M. Vieille, 
Ingenieur des Poudres et Salpetres, Vol. II., 1889; Paris : 
Gauthiers-Villars et Fils. 


the article referred to constituting Appendix I of the 
present work. 

Vieille's method consisted in selecting acid mix- 
tures of specified standard strengths, nitrating under 
regularly controlled conditions as to mass, time, and 
temperature, determining the nitration of- the resul- 
tant products, and plotting upon diagrams the resul- 
tant nitrations as referred to the strengths of the acid 
mixtures used to produce them. In this manner 
he was enabled to recognize a tendency of nitration 
towards groupings or discontinuities, rather than 
towards progressively increasing nitration, advancing 
in measure with increase in strength of acids ; and he 
interpreted this tendency as indicating the existence 
of a number of definite cellulose nitrates. He sum- 
marizes the results of his work as follows : 

"In order to account completely for the different 
changes (groupings, discontinuities) by the production 
of nitro-products corresponding to definite formulae, the 
equivalent of nitro-cellulose must be quadrupled. Nitro- 
celluloses corresponding to such formulae agree with 
the theoretical yields of nitrogen dioxide per gram of 
material indicated in the following table, and which 
correspond either to the discontinuities to which we 
have alluded, or else to a change of physical properties. " 

While Dr. Eder formulated six varieties of nitrated 
material, doubling the coefficients of cellulose by 
writing it C ia H 20 O 10 , instead of C,H 10 O 6 , M. Vieille 
formulates no less than eight compounds, to express 
which he quadruples coefficients, writing cellulose as 
C a4 H 40 O a .. He thus obtains: 



rt w 

in U"> OJ N 'T CO N O* 

MM Qs CO VC Tt f> O 

N <N MM M M M M 

H o 















































o o 

K K 

o o 
X fc 


o o 


K ffi 

<J u 

q q c q 

ffi ffi E 

O U U U 


To illustrate the accordance of theory with prac- 
tice, the content of nitrogen (dioxide), as it should 
exist according to theory, is compared in each case 
with that actually determined from practical experi- 
ment. From the above there would appear to be 
two varieties of gun-cotton insoluble in ether-alcohol, 
two varieties of soluble nitro-celluloses, and a number 
of varieties of nitro-celluloses of lower nitration. M. 
Vieille's paper was published shortly after the official 
announcement by the French Government of the de- 
velopment and successful establishment of the manu^ 
facture of an efficient smokeless powder, these results, 
being the declared fruits of M. Vieille's researches, 
and it is therefore authoritative. 

The points to which attention are especially called 
in M. Vieille's work are the quadrupling of the expo- 
nents of cellulose, and the formulation of as many as 
eight varieties of its nitro-derivatives. 

The abandonment of the old types of smoke-form- 
ing powders that had been in use for hundreds of 
years, and the substitution therefor, by the French 
government, of a new and efficient smokeless powder, 
was a step that naturally attracted the attention of all 
civilized powers. The composition of the French 
powders and their methods of manufacture remained 
carefully guarded secrets. Efforts to achieve similar 
results were at once inaugurated in other countries, 
and the period of inactivity following the abandon- 
ment of efforts to employ gun-cotton as a propellant 
gave way to one of marked activity in all that related 
to the study of explosives. Shortly after the an- 


nouncement of the results obtained in France, the 
Russian government commissioned Professor D. Men- 
dele*ef, a chemist who had already achieved world- 
wide reputation as the expounder of the Periodic 
Law of existence of the elements, to conduct a series 
of researches with a view to the production of an 
efficient smokeless powder for Russia. As the result 
of his labors Professor Mendeleef succeeded in de- 
veloping a powder called by him " pyrocollodion," 
which proved satisfactory, and which was adopted 
in Russia for use in arms of all calibres. As in the 
case of its predecessor in France, the composition and 
method of manufacture of pyrocollodion remain care- 
fully-guarded national secrets. Outlines of ballistic 
results have been published, however, and it is by 
these that the next light is thrown upon the composi- 
tion of nitro-celluloses. 

The scope of Professor Mendeleef's work, and the 
character of the analytical methods followed by him, 
may be ascertained from an examination of a paper 
published by him, my translation of which constitutes 
Appendix II of the present work. In relation to 
the structure of the nitro-celluloses he states : 

" In all aldehydes, beginning with the formic and 
the acetic, a tendency towards polymerization is to be 
noted, due, doubtless, to the property of aldehydes 
of entering into various combinations (with H 2 O, 
NaHSO 3 , etc.); whence the composition C 6 H, O B , 
containing an aldehyde grouping, should also possess 
this property, so far as relates thereto. We may 
therefore safely assume that the molecular composi- 


tion of cellulose, judging from its properties, is poly- 
merized, i.e., it is of the form C 6M H 10M O 6M , where n is 
probably very great." 

It will be seen from the foregoing that in their ef- 
forts to explain the atomic constitution of nitro-cellu- 
lose, investigators have invariably been led to increase 
the common multiple of the elements entering into 
the composition of cellulose, until finally Mendele"ef 
states as his opinion that this multiple may be very 
great through reason of the presence of certain ten- 
dencies towards polymerization. It is interesting to 
observe how different chemists have been led to es- 
tablish conclusions as to the existence of compounds 
corresponding to the formulae they have written. 
Eder dissolves out of gun-cotton its soluble content, 
and as the result of the analysis of the remaining por- 
tion writes its formula as double tri-, or hexanitro- 
cellulose, C 12 H 14 (NO 2 ) 6 O 10 ; he isolates from cotton 
treated with weaker acids, a compound corresponding 
in content of nitrogen to what a body of constitution 
C la H 16 (NO 3 ) 6 O 10 , a pentanitro-cellulose, should give, 
and obtains through the agency of solution a nitro- 
cellulose precipitate satisfying the desired conditions. 
He also obtains, by employing acids of still further 
reduced strengths, a body of variable composition 
which he regards as a mixture of tetra- and trinitro- 
celluloses, but does not succeed in isolating the two 
distinct bodies from the mixture. Vieille employs 
mixtures of acids increasing in strength progressively ; 
continues the process of nitration through a period of 
time sufficiently long to insure the total conversion of 


the cellulose into nitro-cellulose (as shown by the em- 
ployment of a solution of iodine in potassium iodide 
as an indicator of free cellulose); refers to coordinate 
axes the results of each observation, employing 
strengths of acids as abscissae, and numbers of cubic 
centimetres of nitric oxide evolved as ordinates. 
Upon comparing results he notes a tendency towards 
the existence of progressive steps in nitration; i.e., 
for an acid varying somewhat above and below a cer- 
tain point in strength there appears to be formed a 
definite product of nitration. Taking into account the 
number of such steps observed and their distances 
apart, he formulates the series of compounds enumer- 
ated in a preceding paragraph. 



CELLULOSE is distinguishable from other materials 
employed for purposes of nitration by the possession 
of a continuous, complex, and definite cell struc- 
ture. The plant producing it has developed by 
growth from the protoplasmic state, through influences 
of soil, atmosphere, and sunlight; after its death the 
cellulose tissues remain, the skeleton of the once liv- 
ing organism, possessing a structure bestowed by suc- 
cessive growth processes and not by any definite and 
general chemical change occurring at any stated time. 
It is this remaining tissue, an organization of cells 
infinite in their variety and form, that constitutes the 
base material employed for nitration. We may pro- 
ceed to nitrate it in various ways ; we may first par- 
tially destroy or disintegrate the cell structure (con- 
vert it into hydrocellulose) ; the nitration may be con- 
ducted at various temperatures and continued for 
different lengths of time with the employment of vari- 
ous acid mixtures of different strengths, and with the 
use of different relative quantities of cotton and acids. 
If all of these governing conditions be taken account 
of, and if proper allowance be made for their effect, 



we may predicate for the nitrated product an exact 
composition and exact qualities; if any of them be 
overlooked, we lose control of physical character and 
chemical composition of the final product. 

The method of accounting for the composition of 
nitro-cellulose by assuming it to consist of a mixture 
of the different members of a graded series of cellulose 
nitrates themselves definite chemical compounds 
was a rational procedure in accordance with established 
chemical usage; yet there existed grave difficulties in 
the way of maintaining such views. In the first 
place, the members of the series could not be sepa- 
rated from one another. Eder was unable to separate 
cellulose tetra- and trinitrates. The impossibility of 
resolving a given nitro-cellulose into definite quantities 
of the various compounds formulated by Vieille, Eder, 
or their predecessors, is generally recognized by every 
investigator to-day. In the second place, the concep- 
tion of a definite period of time being required to ef- 
fect nitration is inseparably connected with the forma- 
tion of these bodies. Thus properly nitrated gun- 
cotton consists of a large quantity of insoluble with a 
small quantity of soluble and a very small quantity of 
unnitrated cotton ; but if nitration be arrested before 
completion, there results a mixture of different 
quantities of nitrated and unnitrated cotton. 

That the nitration of cellulose is a gradual progres- 
sive process, advancing from incipiency through lapse 
of time towards completion, was the next theory ad- 
vanced. This change in treatment of the problem, 
which virtually led to the abandonment of the old 


simple formulation advocated, is discussed with great 
clearness by M. Bruley, a French Government chem- 
ist, in a paper entitled " Sur la Fabrication des Cotons 
Nitres," published in the Memorial des Poudres et 
Salpetres (Vol. VIII, 1895-6), my translation of which 
constitutes Apendix III of the present work. 

For purposes of elucidation, Bruley's work will be 
considered comparatively, in relation to what has been 
accomplished by others. The first investigators 
treated cellulose with nitric acid direct, and obtained 
a product to which they endeavored to ascribe a for- 
mula; their successors demonstrated the existence of 
more than one variety of nitro-cellulose, and formulated 
a series of compounds. The various acid mixtures 
employed in these researches may be divided into two 
classes: those composed of nitric acid and water taken 
in various proportions, and those containing sulphuric 
acid in additiori to nitric acid and water. It was un- 
derstood how the absorptive action of the sulphuric 
acid rendered possible the formation of cellulose ni- 
trates of higher nitration than those that could be ob- 
tained by the employment of concentrated nitric acid 
alone. But mixtures of concentrated nitric and sul- 
phuric acids necessarily contain water, anhydrous sul- 
phuric acid being a solid and anhydrous nitric acid 
being impossible to prepare ; while with the more 
highly diluted mixture of the two acids cellulose 
nitrates identical with those formed with nitric acid 
and water alone could be formed ; therefore the widest 
range of nitration resulted from the employment of 


mixtures containing nitric and sulphuric acids and 

Vieille had already studied the nitro-products 
formed by the use of mixtures containing various pro- 
portionate quantities of nitric and sulphuric acids of 
definite strengths. Bruley's investigations covered the 
broader field resulting from the employment of mix- 
tures of the three elements, in which the quantity pres- 
ent of each element varied in all practicable proportions 
in reference to the quantities of the other two. The 
scope of the latter work may be graphically indicated 
as follows : 



Let OX and OY be coordinate axes. On OY lay 
off OY' t which divide into one hundred equal parts; 
and let OY" represent the number of parts of nitric 
acid to one hundred parts of sulphuric acid. Then 
Y"X"', parallel to OX, will be the locus of all points 
corresponding to mixtures in which the amounts of 
nitric and sulphuric acids present bear the definite 

ratio 7 to each other. On OX lay off OX' and 

divide it into one hundred equal parts, and let OX" 
represent the number of parts of water to one hundred 
parts of sulphuric acid. Then X" Y'" , parallel to OY, 
will be the locus of all points corresponding to mix- 
tures in which the quantities of water and sulphuric acid 

present bear the definite ratio ~, to one another. 

The point P, in which the lines Y"X'" and X" Y'"' 
intersect, corresponds to a mixture containing definite 
quantities of nitric acid, sulphuric acid, and water. 
The area OY'OX' is a rectangle, every point of which 
corresponds to some one combination of the three 
elements. M. Bruley explores this area by choosing 
a number of approximately equidistant points dis- 
tributed over its surface, preparing the acid mixtures 
corresponding to them, and determining the nitration 
and solubility of the nitro-celluloses prepared from 
these mixtures. Proceeding in this manner, and join- 
ing points of equal nitration, he maps out a series of 
parallel or nearly parallel curves, between which are 
included areas of equal or similar solubility. The fol- 



lowing diagram, taken from M. Bruley's work, illus- 
trates the method : 














5 1015202530354045 


The Roman numeral is the serial number of experiment; the 
upper Arabic numeral, the nitration in cubic centimetres of NO a , 
the second, the solubility; and the third, when given, the vis- 

The nearer the points corresponding to nitrations 
lie to the line Ef> which is the locus of mixtures con- 


taining the least possible quantities of water, the 
higher the nitrations. The area is divided into belts 
corresponding to mixtures forming insoluble nitro- 
celluloses, " intermediate " nitro-celluloses, collodions 
of higher and of lower nitration, and ' ' friable " cottons. 
The published investigations do not consider mixtures 
containing more than 55 parts of nitric to 100 parts of 
sulphuric acid, M. Bruley stating that mixtures con- 
taining more than this relative proportion of nitric 
acid are " not practicable commercially from their in- 
creased cost " ; nor those containing less than 15 parts 
of nitric to 100 parts of the sulphuric acid present; 
in these nitration proceeds with exceeding slowness. 
Similarly, mixtures containing less than 35 per cent, 
of water, compared with the quantity of sulphuric acid 
present, embrace all those capable of producing nitra- 
tions higher than those of "friable" cottons; while 
the lower limit of water is fixed by the strength of 
the strongest acids commercially obtainable. 

Besides nitration and solubility, a new characteristic 
of the nitrated product is here introduced, viscosity, 
as determined by the rate of flow, in drops per min- 
ute, through a standard orifice, of a standard solution 
of the collodion under examination. M. Bruley 
states that increased temperature of nitration, as well 
as continued pulping and washing in warm water, 
all have the effect of diminishing the viscosity of the 
collodions formed from nitro-cellulose ; but he does 
not discuss the practical bearing of this fact on the 
manufacture of colloids. 

M. Bruley also discusses briefly relations of time 


and temperature to nitration. The duration of the 
reaction is found to be the more prolonged the smaller 
the quantity of nitric acid present in the nitrating mix- 
ture; an immersion of two hours is found to suffice, 
in general, for the production of the soluble nitro- 
celluloses ; but as much as eight to ten hours are re- 
quired to complete the nitration of gun-cotton. A 
comparison is made of nitrations conducted at three 
different temperatures, employing the same acids and 
cotton ; and the conclusion is reached that, in the 
case of soluble nitro-celluloses, increase of tempera- 
ture during dipping and reaction increases ultimate 
nitration and solubility; while for gun-cotton, though 
the effect of temperature upon solubility is less dis- 
tinctly marked, yet high temperatures have the effect 
of increasing solubilities. 

During the years 1895 and 1896, I conducted series 
of experiments at the Naval Torpedo Station, at New- 
port, Rhode Island, with the view to the production 
of a nitro-cellulose base suitable for conversion, by 
direct colloidization, into an efficient smokeless pow- 
der.* The problem presented itself in the form of an 
attempt to transform well-known types of nitro-cellu- 
loses, with or without the addition of solid, non- 
colloidable ingredients, and by the use of standard sol- 
vents, into colloid powders; and it ultimately resolved 
into an effort to overcome certain ballistic inconven- 
iences, the existence of which was not recognized when 
the experiments were begun. In the endeavor to over- 

*I have continued experimenting in this field, at intervals, 
ever since this time. J. B. B. 


come these inconveniences, it became apparent that 
there was little hope of preparing from the old mate- 
rials a powder capable of fulfilling service require- 
ments; and the results, wholly negative, of elaborate 
and exhaustive series of experiments with the old 
materials, forced me to the conclusion that, if a suit- 
able powder were developed, it could only be through 
the discovery of a new form of nitro-cellulose capable 
of colloidization, and which would possess physical 
and chemical properties not pertaining to any hitherto 
existing known form of this substance. 

As the result of my labors I succeeded in develop- 
ing such a new form of nitro-cellulose, which I was 
able to convert directly, by colloidization, without 
addition of other ingredient, into a colloid smokeless 
powder, suitable for arms of all calibres. With this 
powder I conducted extended series of experiments, 
testing and establishing its keeping qualities, and stand- 
ardizing weights of charge and dimensions of grain for 
the different calibres of guns. Its manufacture was 
established on a commercial scale at the Torpedo Sta- 
tion, at Newport, Rhode Island, where the represen- 
tatives of the various private powder-manufacturing 
establishments were instructed in the method of prep- 
aration of the new material.* Subsequently these firms 
established plants of their own on a scale far larger 

* At this time (1894-1897) the Naval Torpedo Station, at New- 
port, Rhode Island, was under the command of Captain George 
A. Converse, U. S. Navy, to whose ability as an administrator, 
as well as mechanical skill as a specialist, the successful devel- 
opment of smokeless powder is largely attributable. 


than that undertaken at Newport ; and finally an ap- 
propriation was secured for the Navy, sufficient for the 
erection of a plant large enough to enable this branch 
of the service to manufacture an appreciable share of 
the powder consumed afloat. 

The steps leading up to the development of the 
new nitro-cellulose are cited here in relation to the 
light they throw upon ultimate cellulose and nitro- 
cellulose structure. Taken in order of approximate 
sequence they may be outlined as follows: 

i. Powders were prepared by colloiding insoluble, 
soluble, and mixtures of insoluble and soluble nitro- 
celluloses in acetone (both varieties of nitro-cellulose 
readily colloid in this solvent), forming the resultant 
material into strips of different thicknesses, firing 
charges of different weights of each thickness from 
standard guns, and tabulating resultant velocities and 
pressures. Besides the pure colloids thus obtained 
there were experimented with other colloids having in- 
corporated into them various percentages of barium and 
potassium nitrates, both deposited from solution and 
employed in the dry pulverulent state. It was observed 
that the addition of the nitrates led to the develop- 
ment of smoke, but increased the value of V/P, the 
ratio of the maximum pressure to the initial velocity 

2. A large quantity of gun-cotton, then stored at 
the Torpedo Station, was available for conversion into 
powder. It was found that different manufactured 
lots of this material produced powders differing widely 
in ballistic properties. To determine the cause of 


these differences, I took blocks of each gun-cotton in 
store (some fifty odd lots), washed, dried, and sifted 
each sample, the washing to remove the sodium car- 
bonate, colloided each with the same acetone, formed 
the resultant masses into small rectangular grains of 
the same dimensions and dried them. I then placed 
one- gram weights of each small lot upon slips of glass 
and ignited them in the open. The ashes remaining 
after combustion differed very greatly among them- 
selves, some assuming the form of a fine white resi- 
due, others of a sooty mass of unconsumed carbon. 
Upon comparing the ash residues with the results of 
the complete chemical analyses of the original cottons 
subsequently made, I found that, for the cases where 
carbon was deposited, the gun-cottons from which the 
powder had been made contained an abnormally large 
quantity of free unnitrated cotton, the existence of 
which had not hitherto been suspected. 

3. Those lots of gun-cotton containing large quanti- 
ties of free unnitrated cotton being eliminated, experi- 
ments with powders prepared from the remaining lots 
were continued. Ballistic differences were still found to 
obtain, although less marked than those primarily ob- 
served. Upon comparing the nitration of the original 
gun-cottons, which differed among themselves con- 
siderably, ranging between N = 13.4 and N = 12.4, 
with the ballistic results from the powders prepared from 
them, it was found that the brusquer powders, which 
developed high pressures as corresponding to compar- 
atively low velocities, were prepared from gun-cottons 
of highest nitration. It occurred to me, therefore, 


that uniformity in ballistic conditions might be pre- 
served through the maintenance of mean nitration, 
i.e., by blending different lots of nitro-celluloses to- 
gether, so that the mean nitrogen contents of all 
blends should be maintained a constant ; at the same 
time providing that dimensions of powder-grains and 
weights of charge remained equal. The truth of this 
theoretical assumption was abundantly sustained by 
the results of practical experiment, and the problem 
of the attainment of ballistic uniformity, using gun- 
cotton and the old forms of soluble nitro-celluloses as 
basis materials for the manufacture of colloid powder, 
was solved. Bearing in mind the actual nitrations of 
the various lots of gun-cottons on hand, I chose a 
mean of 12.75 P er cent. N as a standard of nitration, 
and blended the successive lots to this figure, adding 
soluble nitro-cellulose when it became necessary to re- 
duce lots of especially high nitration down to the stan- 

4. I observed, as the result of experiment, that the 
velocity corresponding to a given bore-pressure could 
be increased by incorporating oxygen-carriers, such as 
barium and potassium nitrates, in certain percentages 
into colloid powders; and to such an extent that for 
a pressure of fifteen tons a gain of about 300 foot- 
seconds could be realized. I attributed this difference 
in ballistic effect to an accelerative action due to the 
oxidation of the products of combustion of the nitro- 
cellulose after their evolution in the bore of the gun 
and before leaving the gun, by the gaseous oxides of 


nitrogen evolved from the metallic nitrates.* As the 
bodies to which accelerative effect was attributable 
were free, inert particles distributed throughout the 
substance of the colloid, I argued that free, uncol- 
loided gun-cotton distributed throughout an ether- 
alcohol colloid of soluble nitro-cellulose would produce 
the same effect. Experiments with free gun-cotton as 
an " accelerator" proved failures, however, practically 
no ballistic difference being observed between the ac- 
tion of the ether-alcohol colloid of the soluble, contain- 
ing free insoluble nitro-cellulose and certain other col- 
loids of the same in which both ingredients were present 
in the colloid state. The fact was that both the sol- 
uble and insoluble forms of nitro-cellulose were, in 
both cases, consumed away at practically the same rate. 
5. The effort to prepare powders from colloided 
soluble nitro-celluloses into which uncolloided insoluble 
nitro-cellulose was incorporated as an accelerator, led 
to extended series of experiments with ether-alcohol 
colloids. These proved at first difficult to manufacture, 
but, once prepared, were found manifestly superior to 
the acetone colloids, as they were extremely tough 
and well capable of resisting disintegration in the gun 
at the instant of firing, in which regard the brittle 
acetone colloids had not proved altogether satisfac- 
tory. Actual experiments in firing had also shown 
that the greater the quantity of uncolloided (insoluble) 

* See Appendix IV, which is a reprint of my paper entitled 
"Development of Smokeless Powder," published in the Proceed- 
ings of the U. S, Naval Inst., in which the subject of acceleration 
is treated at some length. 


nitrocellulose present in an ether-alcohol colloid 
powder, the brusquer the powder proved, developing 
a higher bore-pressure as corresponding to the muzzle 
velocity realized. But, in order to maintain a standard 
nitration, explosive strength, it was necessary to 
have present always a considerable quantity of the un- 
colloided (insoluble) form of nitro-cellulose. As this 
was objectionable for the above reason, I therefore 
began experimenting, to see how far the nitration of 
the soluble nitro-cellulose could be raised, with the 
view of minimizing the amount of insoluble nitro- 
cellulose that would be required 

6. I also made investigations in another line, and 
conducted firing trials with a series of powders pre- 
pared from nitro-hydrocelluloses. A note upon the 
constitution of this body may be pertinent here. 

When cotton is exposed to the action of hydro- 
chloric acid, in the form of gas or of concentrated 
aqueous solution, it undergoes a change of form and 
composition, being converted into a material known as 
hydrocellulose, the composition of which has been 
variously regarded. Viewed as a substitution product, 
it is formulated as a cellulose hydrate (Girard, cited by 
Cross and Bevan) ; considered physically, it appears to 
consist of an aggregation of fragments of the cells 
from which the original cellulose was built up. The 
fibres seem attacked along planes where their sub-, 
stance is most readily susceptible of decomposition by 
the acid and are separated from one another; the ac- 
tion of the acid appears to be a cutting of cell-joints 
over planes of weakness, whereby the fibres are di- 


vided into small lengths which are clearly visible un- 
der the microscope. (There would thus seem to be 
some portions of the cell more readily susceptible to 
attack than other portions.) If the process be incom- 
plete, partial separation only occurs, long unattacked 
fibres remaining in quantity. 

Samples of nitro-hydrocellulose, wholly soluble in 
ether-alcohol, of nitration as high as 12.6 to 12.7 
per cent., were obtained by nitrating hydrocellulose 
in the acid mixtures I was employing to prepare 
directly mixtures of insoluble and soluble nitro-cellu- 
lose from cotton. This led to extended series of trials 
of powders, of both the accelerated and the unaccel- 
erated types, containing different quantities of the two 
varieties of nitro-hydrocelluloses in various propor- 
tions, with a view of investigating their stability and 
their ballistic properties. 

On account of their extreme brittleness, these col- 
loid powders proved too brusque, and experiments 
with them were abandoned. They also failed in cer- 
tain cases to develop normal stability. 

7. The observation of certain remarkable phenom- 
ena connected with the effect of decrease in tempera- 
ture in promoting the solution and colloidization of 
certain forms of nitro-cellulose in certain solvents, 
to which reference is made in detail in a subsequent 
portion of this work, led me to take up the consider- 
ation of temperature, as an important factor in the con- 
trol of the character and degree of nitration. I con- 
ducted extended series of experiments in which cotton 
was nitrated in various acid mixtures at temperatures 


ranging between o C. and 80 C. The result of these 
experiments was the development of pyrocellulose, a 
form of soluble nitro-cellulose of high nitration, which, 
for a given weight of its substance, converted into 
colloid grains of standard dimensions and dried, de- 
veloped, when fired from the standard gun, the high- 
est muzzle velocity, as compared with a given limiting 
bore pressure. 

The results of the observations in this regard may 
be briefly summarized as follows : 

The physical and chemical, characteristics of prod- 
ucts of nitration are subject to radical modifications, 
through variation of temperature at which the reac- 
tion is conducted. Increase of temperature has the 
effect of raising the nitration of both the soluble and 
the insoluble varieties, and would appear also to in- 
crease the percentage of the soluble component in a 
blend of the soluble and insoluble varieties. The heat 
employed may be derived from two sources: (i) that 
evolved during the exothermic reaction of the nitra- 
tion of cellulose; (2) that which is supplied from ex- 
ternal sources to additionally raise the temperature of 
the nitrating mass. If the temperature of the nitrat- 
ing mass be raised above a certain point, the structure 
of the cellulose is attacked, and the nitro-celluloses 
cease to form, the material dissolving in the acid, with 
the resultant formation of nitro-substitution bodies of 
other genera, such as the nitro-saccharoses; or else, 
direct decomposition of the nitrated body ensues, 
with evolution of copious fumes of nitric-oxide gas. 
The chemical and physical phenomena attending 


both the decomposition and formation of nitre-cellu- 
loses are largely controlled by temperature. Thus, 
the explosive force of gun-cotton is greatly reduced 
by freezing the latter. Gun-cotton saturated with 
liquid air is not only not an explosive, but is practically 
a non-combustible ; while non-nitrated cotton under 
similar conditions is a violent explosive.* 

Again, as will be shown hereafter, the extent of solu- 
bility of certain nitro-celluloses in certain solvents is 
a function of the temperature at which the solution is 
undertaken ; so that the relative quantities of the 
soluble and insoluble constituents in a mixture of 
nitro-celluloses formed at one operation may depend 
upon the temperature at which the separation of the 
sub-constituents is effected. 

* The result of experiments made by me in New York in Octo- 
ber, 1899. 



CELLULOSE is found to possess, after nitration, a re- 
markable property that unnitrated cellulose does not 
possess: it dissolves freely in a number of liquids, in 
which it is not soluble in the unnitrated state. Of these, - 
the solvents ethyl ether, ethyl alcohol, ether-alcohol, 
and acetone are of special interest, both from their 
connection with the manufacture of smokeless powder, 
and from the physical and chemical bearings of their 
methods of effecting solutions. From the solutions 
the nitro-cellulose forms with them it may be precipi- 
tated by the addition of an excess of water or other 
liquid in which it is not soluble, in a flocculent form; 
and ultimately, after drying, forms a pulverulent mass. 
It is to be specially remarked that the nitro-cellulose 
cannot be recovered in its original cellular state ; the 
process of solution has destroyed its organic structure, 
which may not be recovered or recreated. 

The process of effecting the solution of the nitro- 
cellulose may be regarded as preliminary to the forma- 
tion of the colloid. All forms of nitro-cellulose dis- 
solve freely in an excess of those solvents, in contact 
with smaller quantities of which they form colloids 



directly. If the quantity of solvent be reduced below 
a certain point in proportion to the quantity of nitro- 
cellulose employed, colloidization ensues without pre- 
vious liquefaction ; if the solvent be sufficient in 
quantity to effect liquefaction, evaporation of excess 
of solvent must precede colloidization; if the one state 
can be produced, the possibility of forming the other 
may be predicated. 

The line between solution and colloidization is not 
to be drawn sharply ; the two states of matter merging 
into one another. There are, however, two distinct 
sets of progressive steps or series of physical changes 
to be observed, through one or the other of which 
these bodies pass in their transformation from the 
liquid to the solid state, and which may be expressed 
with reference to their progression as follows: 

First series: (a) liquid; (b) jelly ; (c) elastic mass; 
(d) tough colloid. 

Second series : (a) liquid ; (f) slime ; (c] plastic mass ; 
(d) brittle colloid. 

To Series I belong most ether-alcohol colloids; to 
Series II, most acetone colloids. 

The purpose of the present chapter is to describe 
nitro-cellulose solutions, the methods of forming them, 
and their characteristics apart from consideration of 
their colloidal evaporated residues; and to present in' 
relation thereto certain theoretical considerations 
throwing light upon the chemical constitution of cel- 
lulose and its nitrates. The subjects of colloids, their 
properties and methods of formation, will be treated 


A discussion of the solubility of nitro-celluloses may 
best be prefaced by an account of a remarkable prop- 
erty of soluble nitro-celluloses in general, especially 
characteristic of pyrocellulose and of those forms of 
nitro-hydrocellulose soluble in ether-alcohol; viz., 
the direct solubility of these bodies in the solvent ether * 
when subjected to the influence of cold. 

In August, 1896, I conducted a series of experi- 
ments with soluble nitro-hydrocellulose of very high 
nitration, to determine its adaptability for conversion 
into smokeless powder; this material, colloided in 
ether-alcohol and dried, had given promise of value as 
a progressive explosive. 

Upon a very warm Sunday afternoon, I visited a 
dry-house in which a small tray of soluble nitro-hydro- 
cellulose, of high nitration (N 12.4-!-), was exposed 
to a moderate drying temperature. Removing about 
one gram of the amorphous gray powder from the dry- 
ing-tray, I introduced it into a t f est-tube, which I 
tightly closed with a rubber stopper to prevent the 
absorption of moisture. I then visited the chemical 
laboratory and partly filled the tube with what I sup- 
posed to be ether-alcohol. To my disappointment 
the precipitate did not dissolve. Upon agitation, it 
diffused itself throughout the liquid, but rapidly settled 
to the bottom, in its original pulverulent form, when 
brought to rest. I at first attributed this behavior 
to excessive nitration, believing the material to con- 

*When " ether " and " alcohol " are referred to without quali- 
fication, they are intended to designate ethyl ether and ethyl 


sist of insoluble instead of soluble nitro cellulose. Hav- 
ing observed, however, during the previous winter that 
certain of the insoluble nitro-hydrocelluloses appeared 
to exhibit a tendency to enter into solution upon ex- 
posure to cold (a portion of the precipitate appearing 
to rise in the tube, like a jelly, under the influence 
of cold), it occurred to me to try the effect of a salt- 
and-ice freezing mixture upon the tube and its con- 
tents. I therefore mixed a small quantity of salt and 
ice in a beaker, into which I introduced the tube in a 
central vertical position. To my great satisfaction, 
the whole of the precipitate went rapidly into solution, 
forming a yellowish-brown, mobile fluid. 

While considering the phenomenon that I had wit- 
nessed, I remained seated at my desk, holding the 
test-tube enclosed within the palm of my hand. Sud- 
denly I noticed that, although I had shifted the tube 
into the inverted position (cork downwards), yet the 
contents, masked by my hand, had not observed the 
law of liquid flow, for the lower uncovered end of the 
test-tube, which extended downwards, was empty. 
Carefully opening my hand, I was surprised to find 
that the contents of the tube had condensed into a 
dense jelly, which remained fixed in the upper part of 
the tube. I re-introduced the tube into the freezing 
mixture ; its contents liquefied as before, being trans- 
formed into a liquid as mobile as maple-syrup ; re- 
moval from the source of cold and immersion in water 
heated to about 100 F. (38 C.), or holding in the 
palm of the hand, sufficed to cause the liquid to con- 


The phenomenon above described related to the 
behavior of soluble nitro-hydrocellulose in the pres- 
ence of an excess of solvent (ether) when contained in 
a sealed vessel which is exposed successively to differ- 
ent temperatures; removal of the cork with subsequent 
evaporation of the solvent results in the formation of 
a solid colloid residue, non-liquefiable upon variation 
of temperature. 

After producing the results above described, I next 
endeavored to duplicate them. To my surprise, the 
soluble nitro-cellulose, which I supposed the same as 
that I employed the day before, promptly went into 
solution in ether-alcohol at a temperature of about 
70 F. Investigation led to the discovery of an un- 
fortunate interchange of trays in the dry-house, which 
rendered it impossible to repeat with certainty the 
experiment cited. There was, therefore, but one 
mode of procedure left to select sample lots from all 
the trays, one of which must represent the lot origin- 
ally taken, and experiment with them in the presence 
of ether-alcohol. Still I was unable to obtain the 
original results. It next occurred to me that I might 
have employed by mistake some solvent other than 
ether-alcohol, and as the result of a day's experiment- 
ing, I found that phenomena identical with those 
originally developed could be obtained with soluble 
nitro-hydrocelluloses using ethyl ether as a solvent. 

Experimenting subsequently, 1899-1900, to ascer- 
tain whether the above phenomena were peculiar to 
nitro-hydrocelluloses, or whether they were character- 
istic of all soluble nitro-celluloses, I found; 


I. That pyrocellulose could be readily dissolved 
in ether upon application of cold. 

2. That all soluble nitre-celluloses acted similarly 
in presence of an excess of ether, but that some were 
more readily disintegrated by ether upon application of 
cold than others. 

3. That if the necessary degree of cold were de- 
veloped, soluble nitro-celluloses were not only soluble 
to any desired extent in ether, but that they could be 
colloided directly therein, without recourse to lique- 

4. That the addition of a few drops of alcohol in 
difficult cases appeared to be equivalent to a lowering 
of temperature, i.e., it rendered a given soluble nitro- 
cellulose more soluble in ether, in the presence of a 
given degree of cold, than it otherwise would have 

The problem of nitro-cellulose solution may be ap- 
propriately prefaced with the query : Why are certain 
forms of soluble nitro-cellulose readily soluble at ordi- 
nary atmospheric temperatures in a compound of two 
parts by weight of ethyl ether with one part by weight 
of ethyl alcohol; whereas, the said forms of nitro-cel- 
lulose are not soluble to any appreciable extent, at 
ordinary atmospheric temperatures, in an excess of 
either ethyl ether or ethyl alcohol, when either is 
employed alone as a solvent? The point emphasized 
is, why should the single material, soluble nitro-cellu- 
lose, prove more soluble in a mixture of two solvents 
than in either solvent separately? 

If one-tenth of a gram of soluble nitro-cellulose be 


placed in a test-tube and covered with, say, 25 c.c. of 
ethyl alcohol, and the test-tube be corked and then 
violently agitated, it will be found that solution will 
not ensue, but that the soluble nitro-cellulose will (if 
pulped) gradually settle to the bottom after the tube 
is brought to rest, or else remain suspended in an 
undissolved state in the liquid. Similarly, the same 
quantity of nitro-cellulose will remain undissolved in, 
say, twice the same quantity of ether under similar 
treatment. If, however, the contents of the two tubes 
be combined, the soluble nitro-cellulose will promptly 
go into solution in the mixture of the two solvents. 

Solubility in ether and solubility in alcohol must be 
touched upon before proceeding to the question of 
solubility in the compound solvent. It is known that 
certain nitro-celluloses of low nitration are soluble in 
ethyl alcohol. To this Vieille and Mendeleef attest, 
and the latter recommends that this fact be taken 
advantage of to remove traces of nitro-celluloses of 
low nitration from those of higher nitration, the 
higher soluble nitro-celluloses not being soluble in 
alcohol, at least not at ordinary temperatures. 

If it happened that the soluble nitro-cellulose dis- 
solved with the same ease in warm alcohol that it does 
in cold ether, the action of the compound solvent 
could be accounted for on simple physical grounds. 
A mixture of the two solvents might so balance differ- 
ences of temperature as to effect solution at some tem- 
perature that represented a mean proportional to the 
percentage of each solvent in the mixture. The at- 
tempt was made, therefore, to dissolve pyrocellulose 


in alcohol (95 -per cent.) heated to near its boiling 
point, but proved unsuccessful. On the contrary, 
the tendency was rather away from than toward solu- 
tion. This theory was, therefore, untenable. 

Having alluded briefly to the effect upon soluble 
nitro-cellulose of the individual solvents, (i) ethyl 
alcohol, and (2) ethyl ether, the solubility in the com- 
pound ether-alcohol solvent may next be considered. 

In their work, " Cellulose " (London: Longmans, 
Green & Co.), Cross and Bevan, referring to cellulose 
and its hydration, state (p. 1 1) : 

11 According to modern views on the subject of solu- 
tion generally, and the solution of colloids in particu- 
lar, the lines drawn by the older investigators of these 
phenomena are of arbitrary value ; gelatinization being 
expressed as a continuous series of hydrations between 
the extreme conditions of solid on the one side and 
aqueous solution on the other." 

In their physical forms the solutions of nitro-cellu- 
loses in ether and ether-alcohol present striking analo- 
gies to the hydrated and gelatinized forms of cellulose 
itself. If, as stated, the gelatinized or hydrated form 
may be regarded as a continuous series of hydrations 
of cellulose, then the colloids can be regarded as what 
may be termed a continuous series of etherizations or 
alcoholizations of nitro-cellulose. In this manner we 
may establish an analogy in point of behavior between 
cellulose in the presence of water and nitro-cellulose in 
the presence of ether, alcohol, and ether-alcohol. 

Referring again to Cross and Bevan, "Cellulose," 
we find (pp. 4 and 5): 


"All vegetable structures in the air-dry condition 
retain a certain proportion of water, or hygroscopic 
moisture, as it is termed, which is readily driven off at 
I OO (C.), but reabsorbed on exposure to the atmos- 
phere under ordinary atmospheric conditions." 

" The phenomenon is definitely related to the pres- 
ence of OH groups in the cellulose molecule, for in 
proportion as these are suppressed by combination 
(with negative radicles to form the cellulose esters) the 
products exhibit decreasing attractions for atmospheric 
moisture. It is to be noted that some of these syn- 
thetical derivatives are formed with only slight modifi- 
cations of the external or visible structure of the cellu- 
lose, of which, therefore, the phenomenon in question 
is again shown to be independent." 

In parallel to this we may state: 

The series of the cellulose esters known as the nitro- 
celluloses exhibit at ordinary atmospheric tempera- 
tures, with ether and alcohol, effects similar to those 
of cellulose with water in what relates to, (i) hygro- 
copic moisture, and (2) gelatinization. 

In relation to the effect of alkalies on concentrated 
solutions and this is of primary importance in connec- 
tion with our subject Cross and Bevan state (p. 23) : 

" Cold solutions of the alkaline hydrates of a certain 
concentration exert a remarkable effect upon the cellu- 
loses. Solution of sodium hydrate, at strengths ex- 
ceeding 10 per cent. Na a O, when brought into contact 
with the cotton fibre, at the ordinary temperature, in- 
stantly changes its structural features, i.e., from a flat- 
tened riband, with a large central canal, produces a 


thick cylinder with the canal more or less obliterated. 
These effects in the mass, e.g., in cotton cloth, are 
seen in a considerable shrinkage of length and width, 
with corresponding thickening, the fabric becoming 
translucent at the same time. The results are due to 
a definite reaction between the cellulose and the alka- 
line hydrates, in the molecular ratio Ci a H ao O, : 2NaOH, 
accompanied by combination with water (hydration). 
The compound of the cellulose and alkali which is 
formed is decomposed on washing with water, the 
alkali being recovered unchanged, the cellulose ap- 
pearing in a modified form, viz., as the hydrate 
C ia H ao O 10 .H a O. By treatment with alcohol, on the 
other hand, one half of the alkali is removed in solu- 
tion, the reacting groups remaining associated in the 
ratio C, a H ao O, : NaOH. The reaction is known as 
that of mercerization, after the name of Mercer, by 
whom it was discovered and exhaustively investigated. 
Although, however, it aroused a good deal of attention 
at the time of its discovery, it remained for thirty years 
an isolated observation, i.e., practically undeveloped. 
Recently, however, the alkali cellulose has been made 
the starting point of two series of synthetical deriva- 
tives of cellulose, which must be briefly described." 

" From the points established by Mercer in connec- 
tion with this reaction, the following may be further 
noted : 

"At ordinary temperatures a lye of 1.225-1.275 
sp. gr. effects 'mercerization' in a few minutes; 
weaker liquors produce the result on longer exposure, 
the duration of exposure necessary being inversely 


as the concentration. Reduction of temperature pro- 
duces, within certain limits, the same effect as in- 
creased concentration. The addition of zinc oxide 
(hydrate) to the alkaline lye also increases its activity. 
Caustic-soda solution of i.ioo sp. gr., which has only 
a feeble 'mercerizing* action, is rendered active by 
the addition of the oxide in the molecular proportion 
Zn(OH) a :4NaOH." 

Two points in the above merit special attention in 
connection with our subject. They are : 

I. The removal of one-half of the alkali on treat- 
ment with alcohol, the reacting groups remaining as- 
sociated in the ratio C ia H 90 O ]0 : NaOH. 

2. The acceleration, on exposure to a lye of 
1.225-1.275 sp. gr., of the process of " mercerization " 
by reduction of temperature. Here is presented an 
analogy to the increased solubility of nitro-cellulose in 
ether and ether-alcohol upon application of cold. 

In referring to the production of cellulose thio- 
carbonates and to the quantitative regeneration of 
cellulose from solution as thiocarbonate, Cross and 
Bevan state (pp. 29-31): 

4 'The occurrence of this reaction, under what may 
be regarded as the normal conditions, proves the 
presence in cellulose of OH groups of distinctly alco- 
holic function. The product is especially interesting 
as the first instance of the synthesis of a soluble cellu- 
lose derivative i.e., soluble in water by a reaction 
characteristic of the alcohols generally. The actual 
dissolution of the cellulose under this reaction we can- 
not attempt to explain, so long as our views of the 


general phenomena of solution are only hypotheses. 
There is this feature, however, common to all pro- 
cesses hitherto described, for producing an aqueous 
solution of cellulose (i.e., a cellulose derivative), viz., 
that the solvent has a saline character. It appears, 
in fact, that cellulose yields only under the simulta- 
neous strain of acid and basic groups, and therefore 
we may assume that the OH groups in cellulose are 
of similarly opposite function. In the case of the 
zinc-chloride solvents there cannot be any other de- 
termining cause, and the soluble products may be re- 
garded as analogous to the double salts. The reten- 
tion of the zinc oxide by the cellulose, when precipi- 
tated by water, is an additional evidence of the pres- 
ence of acidic OH groups; and conversely, the much 
more rapid action of the zinc chloride in presence of 
hydrochloric acid indicates the basicity of the mole- 
cule, i.e., of certain of its OH groups. On the other 
hand, in both the cuprammonium and thiocarbonate 
processes there may be a disturbance of the oxygen 
equilibrium of the molecule; and although there is no 
evidence that the cellulose regenerated from these 
solutions respectively is oxidized in the one case or 
deoxidized in the other, it is quite possible that tem- 
porary migration of oxygen or hydrogen might be 
determined, and contribute to the hydration and ulti- 
mate solution of the cellulose. But, apart from hy- 
potheses, we may lay stress on the fact that these pro- 
cesses have the common feature of attacking the cellu- 
lose in the two directions corresponding with those of 
electrolytic strain ; and it is on many grounds prob- 


able that the connection will prove casual and not 
merely incidental." * 

It is the feature of double attack upon the cellulose 
that suggests the cause of the increased solubility of 
the nitro-cellulose in the compound ether-alcohol sol- 
vent, as compared with its relative solubility in the 
ether and alcohol separately. In the case of the 
" mercerized " cellulose we observe that a removal of 
one-half the alkali is effected by the alcohol treat- 
ment. The reaction of the alcohol here relates to the 
basic OH groups, the alkali appearing to be retained 
as a base in the groups of acid reaction. A lower 
temperature facilitates the " mercerization " process; 
this may be interpreted into the statement that raising 
the temperature of the molcule beyond a certain point 
retards the process. So, in the case of the increased 
solubility of soluble nitro-cellulose in ethyl alcohol at 
low temperatures, the employment of artifical cold 
accelerates the process of solution. We may here as- 
sume that the material, from its dual character, is 
under an electrolytic strain, from which it is removed 
by application of cold (abstraction of heat) ; the alco- 

* Referring to the character of the regenerated cellulose in 
comparison with the original material, Cross and Bevan state, 
inter al. : 

" (i) Its hygroscopic moisture, or water of condition, is some 3 to 
4 per cent, higher, viz., from 9 to 10.5 per cent. 

" (2) Empirical composition. The mean results of analysis show 
C 43-3 P er cent., H 6.4 per cent., which are expressed by the 
empirical formula 4C 8 HioO6.H 2 O." 

It will be observed that the regenerated material is an amor- 
phous hydrocellulose. 


hoi then attacks and dissolves the basic OH groups; 
the structure of the cellulose is destroyed ; the acid 
groups yield subsequently and enter into solution, to 
effect which there may occur, as suggested by Cross 
and Bevan for cellulose, a temporary transfer, within 
the molecule of nitro-cellulose, of hydrogen and oxy- 
gen ; and the whole substance finally yields to alco- 
holization in a manner analogous to that by which 
cellulose itself yields to hydration. 

Similarly, for the increased solubility of soluble 
nitro-cellulose in ethyl ether at low temperatures, the 
employment of artificial cold accelerates the process 
of solution. The ether dissolves the acid OH groups 
(in contradistinction to the action of the alcohol, 
which attacks the basic OH groups); the structure of 
the cellulose is destroyed ; the basic groups yield sub- 
sequently, to effect which there occurs a transfer of 
hydrogen and oxygen within the molecule in a direc- 
tion opposite to that in which it occurs in the previ- 
ous case, and the whole substance finally yields to 
etherization in a manner similar to that by which cel- 
lulose itself yields to hydration. 

As bearing upon the theory it may be remarked that 
the two solvents, ethyl alcohol and ethyl ether, differ 
by H-O-H, since ether, (C,H 6 ),O, + H a O = alcohol, 
2C,H 6 0. 

In the case of the mixed ether-alcohol solvent we 
subject the cellulose in the one operation to the double 
tendency to disintegration. The necessity for the 
application of cold (or the absorption of heat) to ef- 
fect the necessary arrangement of the H-O-H groups, 


as they may be styled, i.e., to strain them into basic 
and acid relations, is no longer required, for the double 
attack in two directions, corresponding to those of 
electrolytic strain, is provided for by the parallel double 
composition of the ether-alcohol solvent. 

There is reason for supposing that the above-de- 
scribed reaction is not limited in its occurrence to 
forms of soluble nitro-cellulose alone, but that it 
obtains equally for the insoluble varieties. If the 
theory be correct, then insoluble nitro-cellulose should 
be soluble in ether. Macnab has shown that insoluble 
nitro-cellulose dissolves in ether-alcohol at a very low 

In the fall of 1899, the time that liquid air first be- 
came readily obtainable in quantity in New York, I 
decided to check Macnab's experiment by employing 
this material as a refrigerant; also (i) to experiment 
with a view of ascertaining whether insoluble nitro- 
cellulose of very high nitration in its unpulped fibrous 
form, the form in which it might be supposed to 
oppose a maximum resistance to the disintegrating 
action of solvents, would not actually go into solu- 
tion in ethyl ether alone under influence of extreme 
cold ; and (2) to determine how far nitro-cellulose is 
soluble in absolute ethyl alcohol. The ether em- 
ployed in these experiments was Squibbs', C.P., for 

* See experiments cited by Guttmann in his article " Manufac- 
ture of Explosives," published in the Journal of the Society of 
Chemical Industry, London, June 30, 1894 ; also p. 404 appendix v 
to his recently published work, " Manufacture of Explosives," 
London: Macmillan. 


anesthesia; the insoluble nitrocellulose was gun- 
cotton of high nitration (N = 13.4+) and purity, in 
the unpulped state. 

The solvent employed in the alcohol experiments 
was absolute ethyl alcohol ; the nitro-celluloses were 

(1) soluble nitro-hydrocellulose of nitration, N 11.4$; 

(2) soluble nitro-hydrocellulose of N 12. 6$; (3) pulped 
pyrocellulose, N 12. 42$; and (4) unpulped gun-cotton 
of high nitration, N 13.4$. 

Experiment I. I placed about one-tenth gram of 
gun-cotton in a test-tube, poured over it 25 c.c. of 2 :i 
ether-alcohol, tightly closed the tube with a rubber 
stopper and immersed it in a vessel containing about 
750 c.c. of liquid air. When the contents of the tube 
became sufficiently cooled, the gun-cotton went read- 
ily into solution, forming a clear mobile liquid of a 
honey-yellow color. I found that the gun-cotton re- 
mained in the colloid form after the removal of the 
tube from the liquid and the subsequent heating of its 
contents to the temperature of the atmosphere. On 
removing the tube from the liquid-air bath, uncorking 
and allowing the contents to evaporate, the residue 
formed a tough, homogeneous amber-colored film. 

Experiment II. Having succeeded readily in col- 
loiding insoluble nitro-cellulose in ether-alcohol, I next 
decided to ascertain whether it were possible to col- 
loid it in ether alone. I placed one-tenth gram of the 
gun-cotton in a test-tube, poured over it about 25 c.c. 
of ether, tightly closed the tube and immersed it in 
the same quantity of liquid air as before. 

Upon sufficiently reducing the temperature, the 


contents of the tube became solid and the ether froze 
into a snow-white mass. On removing the tube from 
the cold and allowing the ether to melt, I found that 
the gun-cotton had disintegrated and gone into solu- 
tion, forming a mobile slightly clouded liquid with a 
yellowish tinge. The gun-cotton remained in solution 
after the withdrawal of the tube from the liquid air 
and the subsequent heating of its contents to the 
temperature of the atmosphere. On removing the 
tube from the liquid-air bath, uncorking and allowing 
its contents to evaporate, the residue formed a tough, 
homogeneous, slightly yellowish colloid. 

Experiment III. About one-tenth gram of each 
of the above-mentioned series of nitro-celluloses was 
each placed in a separate test-tube and each had 
poured over it about 25 c.c. of absolute ethyl alcohol. 
The tubes thus partly filled were closed tightly with 
rubber stoppers, immersed in liquid air (for about 
two-thirds of their lengths), and allowed to remain 
therein for five minutes.* 

The contents of each tube froze in about half a 
minute and the rest of the time the liquid air was 
acting upon the frozen contents. On removing the 
tube containing the gun-cotton from the bath and 
allowing its contents to melt, the gun-cotton was, to 

* The liquid air for these experiments was very kindly fur- 
nished me by Messrs. Vandivert and Gardenhire, of 32 Broadway, 
New York, pioneers in the development of liquid air in the United 
States. I am also indebted to Mr. M. Burger, president of the 
company, for his kindness; and especially to Mr. O. Ostergren 
of New York,* one of the original experimenters and developers 
of the commercial manufacture of liquid ain 


all appearances, unaltered. This showed that, whether 
or not the fibre of the cotton had been attacked, it 
was not affected to the same extent by the ethyl 
alcohol as it was by the ethyl ether, which, as pre- 
viously shown, destroyed the fibre of the gun-cotton, 
causing it to go into solution. 

The sample of pyrocellulose and the two small lots of 
nitro-hydrocellulose in the other tubes were found to 
have gone into solution in the ethyl alcohol, through 
the action of the cold, the former affording a jelly- 
like straw-colored colloid ; the two latter, the usual 
brownish-yellow colloids. 

We may now call attention in parallel to the follow- 
ing remarkable reactions ; 

I. That nitro-celluloses dissolve in ethyl ether 
under the influence of intense cold. 

2. That nitro-celluloses, with the apparent excep- 
tion of the highly nitrated insoluble variety (this will 
be further experimented with later, in the presence of 
extreme cold) dissolve in ethyl alcohol under the in- 
fluence of intense cold. 

3. That the so-called soluble forms of nitro-cellu- 
lose of mean and low nitrations go into solution in a 
mixture of ethyl ether and ethyl alcohol at ordinary 
atmospheric temperatures. 

In connection with the above may be mentioned : 

(a) That the various forms of nitro-celluloses seem 
more readily soluble in ethyl ether than in ethyl alcohol. 

(b That a mixture of two volumes of ethyl ether 
to one of ethyl alcohol seems to extract at ordinary 
atmospheric temperatures the greatest quantity of 


soluble constituents from a mixture of the two forms 
of nitro-cellulose (soluble and insoluble as formed by 
one dipping of cotton.) 

It now remains to consider the theory of compo- 
sition of cellulose and nitro-cellulose from the chemical 
standpoint. In this connection it is necessary, first of 
all, to emphasize the fact of the similarity in compo- 
sition of the cellulose and the nitro-cellulose molecule. 
The latter is formed from the former by the substitu- 
tion, for a certain number of replaceable hydrogen 
atoms, of the same number of equivalents of nitryl 
NO,. The effect of the introduction of nitryl into the 
substance of the cellulose is an apparent weakening 
of the stability of the latter, rendering it susceptible 
of decomposition in a number of additional ways, and 
making actual tendencies to decomposition, traces of 
which exist in the original mate-rial, but which are held 
in control by stronger counteracting tendencies derived 
from other sources. 

For the sake of simplicity the formulae presented in 
what follows will be those of cellulose. It is to be 
understood that whenever questions arise referring 
directly to nitro-cellulose, such as solubility in ether 
or ether-alcohol, the cellulose molecule as presented is 
to be considered as tacitly representing nitro-cellulose, 
without actual expression of the substitution of NO, 
in the replaceable hydrogen atoms. 

Bearing the above in mind, the following additional 
facts may be taken advantage of in formulating a 
theory of the composition of the cellulose and the 
nitro-cellulose molecule ; 


I. The two solvents of soluble nitro-cellulose, ethyl 
ether and ethyl alcohol, differ in composition by H-O 
H. Like H a O, they may be written (C a H 5 ) a O and 
C,H.O or C a H B .O.H. 

2. Cellulose, C ia H ao O 10 , is transformed by treatment 
with alkali in aqueous solution into C ia H 20 O 10 .2NaOH. 

3. By subsequent treatment with water it is con- 
verted into cellulose hydrate, C ia H 90 .O 10 .H a O. 

4. By treating C^HjD^NaOH with alcohol, 
C a H 6 O, the former is converted into the mercerized 
form, C,,H ao O 10 .NaOH, with half the alkali removed. 

To account for the occurrence of reactions (2) and 
(3), the composition of the cellulose molecule is 
regarded as doubled, i.e., raised from C,H, O 6 to 
C, a H ao O 10 . (4) is of especial importance and interest, as 
it exhibits the action of alcohol upon non-nitrated cel- 
lulose; establishing a basis for the assumption that the 
action of the double ether-alcohol solvent upon nitro- 
cellulose is based upon the original composition of 
cellulose itself. 

Assuming a duality of composition as indicated (i) 
by its basic and acid reactions; (2) by its greater solu- 
bility in the compound solvent; (3) by the H-O-H 
difference of the components of the original compound 
s6lvent ; and (4) by the creation and absorption of the 
H-O-H groups in the formation of cellulose hy- 
drates and alkaline compounds, we may proceed as 
follows : 

If alcohol is capable of effecting the solution of a 
nitro-cellulose, it must effect the solution of both its 
components or resolve itself into two sub-components, 


each reacting on one sub-component of the nitro-cel- 
lulose, in order that it may effect the solution of the 
whole. Under the latter assumption, we may con- 
sider alcohol, C,H 6 O, as having the composition 

H H 

| H H/ | 

H H 

which may be divided into 

H H 


H H' | 

H H 

Similarly, we may regard ethyl ether, (C 2 H B ),O, as 
having the composition 

H H 

I I 

H C O C H 

I I 


I I 

H H 

capable of dividing into 

H H 

H C O C-H 

| and | 

H C H H C-H 


If such a tendency as illustrated above exists, it 
will become a reality through the disintegration of 
equivalents which, in combination, correspond to a 
molecule of water, H 9 O.* The removal of one-half of 
the alkali from cellulose in the mercerization process 
is effected by the action of the alcohol upon the 
basic groups therein ; the molecule of cellulose should, 
therefore, be represented so as to permit a resolution 
of the original material into acid and basic sub-com- 
ponents through the disintegration of the water- 
groups, Under this assumption, we may write cellu^ 
lose as 

H H 

I I 

H C O H H O C H 

H- I 
C O H H C 


C O H H O C 

I -o^ I 

H C O--H H O C H 

H H 

* It will be observed that the constituents of water, H-O-H, 
form the central linking group in each of the expressions 

H H 

H H || 

H C O C H 

and | | 


H H 

H H 


which would resolve into 

H H 

H C O H H O C H 

I /H | 

c< o=c 

I X H | 

C O H H O C 
Form I. |i || 

C O H H O C 

c=o c 

I H/l 

H C O H H O C H 


H H 

I I 

H C O H H O C H 

I /H | 

c< o=c 

|\H | 

C O H H O C 
Form II. || II 

C O H H O C 


H C O H H O C H 

I I 

H H 

The molecule as above written contains double cen- 
tral carbon bonds, which fact permits it, on its enter- 
ing into combination, to be written as 


H H 

H C O H H O C H 


| -H H | 

C O H H O C 

_C O H H O C 

I ^-D-- I 

H C O H H O C 

i ' 


Without radical modification, it may be expressed 
H C O H H O C H 


H C O H H O C H 
H C 0-H H O C 


^ ^ 

I I 

as corresponding to C,,H, O 10 . 


Halving it, we will obtain 

corresponding to C,H 10 O 6 .* 

This latter form, which is the simplest expression 
for cellulose, represents, not the molecule, but the type 
unit of cellulose, as it enters into combination, 
through its four free single carbon bonds, either with 
other similar units, by polymerization, or with other 
substances by chemical combination. 

* Written singly, in a similar form, but with closed bonds, the 
single molecule, CaHioOe, may be expressed as 

H H 

Compare also with Cross and Bevan, " Cellulose," p. 38, where, 
in reference to cellulose acetates, it is stated if the above formula 
(alluding to a formula cited) be established by further and ex- 
haustive investigation, the cellulose unit must be C6H 6 O.(OH) 4 . 


The multiplicity of the cellulose derivatives and the 
generally recognized tendency towards polymerization 
already alluded to suggest the further amplification 
of the molecule by inter-combination of its units 
through connection of their carbon bonds into poly- 
meric forms, the type of which may be expressed as 
follows : 

Such a method of representation possesses a special 
interest, as it prepares the way for other considerations. 
The actual number of phases for the polymerized 
molecule thus expressed may vary from 2 to any 
desired number. In the above diagram a 5-phase 
form is presented for purposes of illustration. The 
simple molecule with its four free carbon bonds as 

6 4 


presented in each of the five sectors combines with 
the simple molecule in each adjacent sector on either 
hand. The simplest polymerized form that, under our 
theory, can stand alone, is C ]2 H ao O 10 , composed of 
two "sectors," the carbon bonds in each of which 
unite with those of the other. 
Thus we have 

correcting the form previously written, 



The 5-phase molecule, or its polymer, corresponds 
to pyrocellulose. 

It is evident that under such an assumption the 
molecule may possess an infinity of phases. There is 
no limit to their number. On this assumption, and it 


seems to me on this assumption only, may we account 
for definite chemical composition of the cellular form 
in the plant structure. For we may regard the cell 
as built up from an aggregate of molecules of identical 
composition but of progressively-varying numerical 
phase. The cell may begin with molecules of low 
phase and end with molecules of high phase, or con- 
versely. Molecules of progressively-varying phase- 
magnitude may be deposited in turn from protoplas^ 
mic matter in particular forms of different density, the 
successive evolutions providing for the infinity of cel- 
lular structures appreciable to the human eye, which, 
in their successive deposition, build up the fibrous sub- 
stance of the cell. 

It is to be remembered that the graphic represen- 
tation of molecular arrangement upon a plane surface, 
space of two dimensions, cannot be regarded as 
more than a conventional device illustrating an ar- 
rangement that exists in nature in space of three 
dimensions. Nevertheless, the conventional ring- 
formed combination of elemental particles shown in 
the polyphase molecule strongly suggests the vortex- 
ring theory of the composition of matter (as applica- 
ble to the molecule]. For whatever the atom may be, 
the molecule need not be limited in composition to 
the simplest collection of the lowest possible mimber of 
atoms capable of entering into combination, but may 
be built up from a very great number of the elemental 
particles taken together in their proper ratios.* Such 

* On this hypothesis the atomic weight of cellulose would be 
represented by an average. 


a molecule would increase in amplitude according to 
the number of elemental particles entering into its 
composition ; and the thought therefore suggests it- 
self, that progressive variation in the amplitude of the 
molecular ring is a characteristic of organic life. Or, 
conversely, we may state that we may seek for the 
beginnings of organic life, at least of plant-life, in 
the polymerization of the carbohydrates. 

Briefly summing up, the conditions governing the 
formulation of cellulose may be stated as follows : 

I. The capability of the expression of the molecu- 
lar formula as (C 6 H 10 O 6 ) or C eM H 10tt O 6M justifies the as- 
sumption that the molecule may be represented as 
composed of n number of similar atomic aggregations 
ol the form C e H 10 O 6 , these aggregations conjointly 
forming the molecule. 

2. Under the assumption that Eder's nitrates repre- 
sent limits of nitration in the sense defined by Vieille, 
n cannot possess a value less than 2 ; that is, C ia H ao O lt 
represents the lowest expression for the molecule. 

The fact that 2 : i ether-alcohol dissolves certain 
forms of nitro-cellulose at ordinary atmospheric tem- 
peratures with greater ease than any other compound 
solvent containing ether-alcohol in other than the 2 : I 
proportion, tends to show that all the OH groups in 
which nitro-substitution takes place are not similarly 
placed within the molecule. 

4. The fact that, at very low temperatures, either 
ether or alcohol singly will dissolve soluble nitro-cellu- 
lose, whereas at ordinary atmospheric temperatures 
a mixture of the two solvents is required to effect 


solution, implies that, under influence of cold (absence 
of heat), both ether and alcohol, on the one hand, 
nitro-cellulose on the other, may undergo a certain 
atomic rearrangement within the molecule. 

5. If symmetry of arrangement exists in the sub- 
components of the molecular section C 6 H 10 O, per- 
mitting the representation of atomic aggregations 
which, according to the influences to which they are 
exposed, may exhibit acid actions on the one hand, 
basic reactions on the other, it would be by groupings 
of the carbon and hydrogen atoms, which are even 
in number, around the oxygen atoms, which are odd 
in number. 

6. There is good reason for the assumption (basis of 
theory of nitro-substitution) that the molecules of both 
cellulose and nitro-cellulose are of similar structure; 
that there is no general rearrangement of the atoms of 
the cellulose in the process of nitration, but that nitra- 
tion is accomplished through the substitution of nitryl 
for replacable hydrogen in certain hydroxyl groups, the 
said groups retaining, after nitration, the position in 
the molecule that they held before nitration, as already 
stated. It is upon this assumption that we represent 
the molecule of cellulose as typical, both of cellulose 
and nitro-cellulose, and do not represent the substitu- 
tion of the nitryl in the molcule, unless actually refer- 
ring to some specific property of the nitro-cellulose 
distinguishing it from the original cellulose. 

7. Under the assumption (i)of the basic and acid, 
positive and negative, distribution of the atoms within 
the molecule, based upon the conception of the con- 


stitution of the molecule as a double salt, and (2) of 
the action of the compound ether-alcohol solvent upon 
nitro-cellulose at ordinary atmospheric temperatures, 
and of the action of both the ether and alcohol sep- 
arately upon nitro-cellulose at low temperatures, 
we may regard (a) the ethyl ether, (&) the ethyl 
alcohol (which possesses the composition but not the 
atomic arrangement of an ether [methyl]), and (c) 
the nitro-cellulose, which possesses the characteristic 
properties of an ether, as all splitting up into dual 
sub-groups. The manner in which ethyl ether and 
ethyl alcohol could so split up symmetrically is easily 
shown ; regarding nitro-cellulose as an ether, we may 
seek to split it into sub-groups within the molecule 
in a similar manner. 

The three type forms may then be presented as fol- 

lows : 

Ethyl alcohol Ethyl ether 

C.H.O (C.HJ.O 

(Under strain as methyl ether) 

H H 

H H || 

| /H H v | H C O C H 

/ \ i 

H H || 

H H 

(As expressed by its type, cellulose) 

I I 

H C O H H O C H 

H C O H H O C H 


That all nitro-celluloses are soluble in the compound 
ether-alcohol solvent, notwithstanding the wide differ- 
ences in temperatures at which solution is effected, 
leads me to the conclusion that all nitro-cellulose 
molecules are of similar constitution and organization. 
That they are of a dual composition is evinced by 
their more ready solubility in the compound solvent ; 
again the fact that they are all soluble in the single 
solvent ethyl ether leads me to the conclusion that 
the two halves of the dual molecule are, with respect 
to each other, similar, or of similar inverted forms. 
Therefore, on page 60, Form I is the correct diagram 
of the separation of the factors, rather than Form II. 

We have considered briefly in the preceding the 
composition of the molecule in relation to mercer- 
ization, polymerization, nitration, and colloidization. 
It remains to dwell somewhat more fully upon the re- 
actions of nitration and hydration, and to show how 
they are reflected in the modified structure of the 

The type cellulose has been written : 

By changing the relative position of atoms of hydro- 


gen, and without altering the constitution of the 
molecule, it may be written : 

H-C O H H 0-^3 H 

H-C-0 H H 
H-C H H O 

H-C 0-H H O- 

The polymerized molecule is composed of sections of 
the type : 

H _ C _0-H H-O-^-H, 
H _i_0-H H-O-C H 

From an examination of the unit of polymerization it 

H G O H H O C H 

will be seen that the atoms of carbon are connected (i), 
by single bonds with each other; (2), with hydrogen 
bonds; (3), with hydroxyl bonds; and (4), with what 

may be termed " water bonds," of the form \ . 

X H 

Each carbon atom is connected with two other car- 
bon atoms, one on either side. Every third carbon 


atom is connected with the di-valent " water " radicle 
(A). The remaining two of each set of three carbon 
atoms are each connected with one atom of hydrogen 
and one hydroxyl radicle (B). The symmetrical order 
of arrangement of the carbons is BAB BAB 

The free carbon bonds of each unit represent poly- 

The hydroxyl radicles represent nitration, in the 
mean and higher stages. 

The " water" radicles represent mercer ization, and 
nitration in the lower stage. 

The open oxygen bond in the " water " radicle rep- 
resents colloidization by combination of the half mole- 
cule of the nitro-cellulose with the half molecule of 
the solvent. 

Nitration. The process of nitration may here be 
taken up for more extended consideration. The fibrous 
cellulose reflects in its chemical behavior a character- 
istic of plant life, namely, the possibility of altering 
growth conditions through changes of temperature. 
The reactions into which it enters may be varied 
through the variation of the temperature at which the 
said reaction takes place. These changes in rate and 
character of reaction as due to temperature are espe- 
cially characteristic of nitration, mercerization and 
colloidization ; and doubtless, during the original 
growth processes of the plant, affect polymerization. 
The effect of temperature upon nitration is reflected in 
the resultant nitrated product in two ways: (i) in 
raising the degree of nitration in proportion to the 


increase in temperature (within certain limits) ; and (2) 
in increasing and modifying solubilities corresponding 
to given nitrations. 

Until very recently, our actual knowledge of the 
nitration of cellulose may be said to have been con- 
fined to the following facts : 

I. That there was a lower limit of obtainable nitra- 
tion, somewhere between that represented by the 
compounds C 13 H 14 O 4 (OH) 6 (NO 9 ), for which N = 3.80, 
and C ia H 14 O 4 (OH) 4 (NO,),, for which N = 6.76. 

2. That there exists an upper limit of attainable 
nitration, somewhere near that of the hexanitrate of 
Eder, C ia H 14 O 4 (NO 8 ) 6 , for which N = 14.14. 

3. That nitration is progressive between these 
limits ; and that when determined at a given tempera- 
ture, it advances gradually and progressively from the 
lower towards the higher limit, while under certain 
conditions it may be made, similarly, to decrease. 

4. That somewhere between the attainable limits of 
nitration a point is reached (and the point is not con- 
stant so far as relates to degree of nitration) where the 
nitrated product ceases to be soluble at a given mean 
atmospheric temperature in a mixture of ethyl ether 
and ethyl alcohol. This fact was utilized to separate 
the so-called '* soluble" nitrates of celluloses of low 
nitration from the ' ' insoluble ' ' nitrates of high nitra- 

Let us consider the formula for the double-type 
molecule in connection with the six nitrates for- 
mulated by Eder, and let us assume that there has 
been isolated from the substance of the cellulose for 


purposes of experimentation a homogeneous body 


H-C O-H H-O-C H 

H-C O-H H-O-C H. 
H-C O-H H-O-C H 

H-C-O-H H-O-C H 


composed of molecules of the above 2-phase form. 
Then, if there are six nitrates, we could form them 
successively by substituting nitryl for the replaceable 
hydrogen in the " water" and "hydroxyl" radicles of 
the cellulose on one side of the molecule. Substituting 
thus in the above diagram we would have, for the 
highest nitrate 


H-C O NO 2 H-O-C H 

TT-6 O NO 2 H-O-<JMEt 
H-C-O NO 2 H-O-C H 

6-=^o^ J3>6 

H-6 O ;NO 2 H-O t H 


On account of the ready solubility of the lower 
nitrates in ethyl alcohol at ordinary temperatures the 
maximum efficiency of the 2 : I ether-alcohol, and the 
solubility of the highest nitrates in ethyl ether, we 
assume that the replaceable hydrogen atoms in the 
water radicles are replaced by nitryl first, and that 
subsequently the nitro-substitution continues through 
the displacement of the hydrogen in the hydroxyl 
radicles. If we regard alcohol as connected with the 



positive (basic) side of the molecule, and ether with 
the negative (acid) side, we may represent nitration, 
for the higher nitrates (say, the hexanitrate), as follows: 


H-C O NO 2 H-O-C H 

<H S~-NQH> 

H-C O N0 2 H-O-C H 
H-C-O NO 2 H-O-C H 
-H I NOi^i 

O N0 2 H-O-C-H 


Under such conditions, we may formulate the 
nitrates of cellulose for the 2-phase molecule as fol- 


Cellulose hexanitrate, C ia H 14 (NO,) 6 O 10 . 

Cellulose pentanitrate, C 12 O 16 (NO,) 6 O 10 . 

Cellulose tetranitrate, C ia O 16 (NO,) 4 O 10 . 

Cellulose trinitrate, C 31 H 17 (NO 3 ) S O 10 . 

Cellulose dinitrate, C 17 H 18 (NO,),O 10 . 

Cellulose mononitrate, C 12 H 19 (NO 2 )O 10 . 

If we turn from the consideration of the theoretical 
2-phase cellulose above represented to that of actual 
cellulose, we will find that the latter may be resolved 
into cellulose nitrates, not in six, but in a very great 
number of ways. Every molecule may be differently 
affected in the process of nitro-substitution. The 
character and composition of the resultant product may 
vary with strength of acid, its relative quantity in pro- 
portion to the cotton taken (this determines the rate 
of absorption of water formed during the reaction), 


the temperature at which the reaction takes place, and 
the duration of immersion in the acid bath. The 
cotton may be nitrated on the surface of the fibre or 
nitrated wholly throughout ; it may represent a mix- 
ture of soluble and insoluble nitrates or it may possess 
a uniform degree of nitration ; and it may exhibit 
variations (for the same chemical composition) in its 
solubilities in standard solvents; for the successive 
polyphase molecules are differently constituted and 
differently placed, and doubtless afford progressively- 
varying resistances to the effect of the acid bath. 

It will be remembered that the theoretical limiting 
of nitration (14. 14 per cent, nitrogen) was not obtained 
by Eder. Probably the ultimate substitution of nitryl 
for every atom of replaceable hydrogen is a limit that 
may not be attained in practice. Doubtless some of 
the molecules do represent complete replacement, 
while others are considerably removed from complete 
transformation. It will be shown hereafter that by 
colloidization the molecular amplitude of the nitro- 
cellulose is reduced to a mean, but such is not the 
case for the fibrous material. 

Hydration. When, in the attempt to nitrate cellu- 
lose, the strength of the acid mixture is reduced below 
a certain figure, a remarkable phenomenon may be ob- 
served instead of a nitration there may be produced a 
hydration. We may, therefore, assume that from the 
same point of weakness in the molecule there may de- 
velop either a nitro-substitution (nitration), or else an 
atomic rearrangement accompanied by an absorption of 
water (hydration). It has already been stated that ni- 


tration starts, in accordance with our theory, in nitro-sub- 
stitution in the water radicles, producing nitro-celluloses 
soluble in the alcohol (basic) solution. The hydration, 
then, should start in the rearrangement of the parts of 
the molecule adjacent to these water radicles. The 
quantity of water that is to be absorbed or, rather, 
taken up into the substance of the molecule, may, as 
shown by Cross and Bevan in their references to mer- 
cerization, approach the limit C ia H ao O J0 .2H a O ; corre- 
sponding to C ia H ao O 10 .2NaOH ; while the definite hy- 
drate, C ia H ao O, . H a O, containing half as much water, 
and corresponding to C ia H 20 O 10 .NaOH, is readily iso- 
lated. We have, therefore, as definite limits, the in- 
corporation into the molecule of cellulose, C 13 H ao O 10 , 
of water in the two ratios, 2H,O and H a O. 

I / \ I 
Now the radicle C< >C plus H,O becomes 

| \HH/ | 
H C O H H O C H ; and as the water radicle 

i i 

occurs twice in the double molecule C,,H, O, ,we have 
for the full transformation, 

i i 

H C O H H O C H 

H C O H H O C H 

I | + H,0 = 

H C O H H O C H 

H C O H H O C H 



H C O H H O C H 

I I 

H C O H H O C H 

C O H 


H C O H H O C H 

H C O H H O C H 

O H H O C H 

H C O H H O C H 

I I + *H 9 = 

H C O H H O C H 

H C O H H O C H 


H C O H H O C H 

H C O H H O C H 
I I 

H C O H H O C H 
I I 

H C O H H O C H 
I | 

H C O H H O C H 


In considering these forms of modification of the 
molecular structure it should be remembered that the 
molecules of the original fibre are characterized by 
their variations in amplitude. And we know, actually, 
that hydration represents a breaking down of the cell- 
fibre. This breaking down is also illustrated in the 
two diagrams of the modified molecules just presented. 
In the former it will be observed that but one of the 
two water-bonds remains; in the latter, that both 
water radicles have disappeared and that the two 
halves of the molecule appear wholly separated. 

The effect of hydration in breaking down the fibrous 
structure of the cotton and ultimately putting it into 
solution may therefore be explained by the following 
steps : 

I . The fibre exists as an aggregation of cells of vary- 
ing amplitude. The solvent .at first attacks those cells 
that are either the weakest or most exposed, and trans- 

I / \ I 
forms in them some of the water radicles, C^ ^ C, 

\\HH/ \ 

\ I 

into the hydroxyl forms, H C O H H O C H. 

I I 

The result is a partial transformation into C, a H ao O 1(l '- 

H 3 = (C ia H aa O u ). 

2. As hydration proceeds, the attack takes place 
throughout all molecules irrespective of amplitude un- 
til in each molecule one-half of the water radicles are 
transformed into hydroxyl forms, and a material, yet 
fibrous, built up of modified cells of varying ampli- 
tude in each of which one-half of the water radicles 


are transformed into hydroxyl forms, results. This is 
the true hydrocellulose, C ia H ao O 10 .H,O, or, more prop- 
erly, C ia H aa O n . 

3. As hydration proceeds, transformation advances 
towards the total conversion of all the water radicles 
into hydroxyl radicles. This is accompanied by dis- 
integration of the fibre and a tendency to enter into 
solution. The formula shows that with total trans- 
formation of the water radicles into hydroxyl forms 
the molecule splits up into two halves, between which 
there is no chemical union ; therefore actual chemical 
combination of the two halves ceases, and there can be 
no chemical connection between them except such as 
may be represented by electrolytic strain. 

4. As solution becomes actually effected, the fourth 
and last, and perhaps the greatest, change in the series 
occurs. By their intimate contact and admixture the 
dissolved molecules, freed from their organic form of 
aggregation, are reduced to a common amplitude. 
They hereafter constitute amorphous cellulose, and 
may be represented as an aggregate of the form 
C 6 H 10 O 6 . It is probable, however, that the number 
of atomic particles in each molecule still remains ex- 
ceedingly great, as progressive nitration still appears 
to occur for this material. 

5. The transformation of the whole substance of the 
cellulose into the form corresponding to C ia H ao O 10 .2H a O 
represents the dividing line between the chemical and 
the physical aspects of the absorption of water into 
the substance of the molecule. Any greater absorption 
of water than that corresponding to C,,H M O U .2H,O 


pertains to simple solution; any less absorption, to 
chemical change, affecting the structure of the cellulose 

6. The hydrocellulose obtained by the usual pro- 
cesses represents, simultaneously, a combination of 
all the above-described processes. Part of the cellu- 
lose remains wholly unattacked, and, if allowed to 
exist, renders the whole mass what the workmen style 
"woolly"; part is converted into C la H aJ O n , a true 
hydrocellulose, which, inasmuch as its fibre still ex- 
ists, combines to exercise the same effect upon the 
physical constitution of the mass as the unnitrated 
portion ; part is precipitated as wholly hydrated cellu- 
lose of the form C ia H 20 O 10 .2H a O, and part goes into so- 

7. Once the fibrous structure is lost by solution, 
and the molecules reduced to a common amplitude, 
the organic constitution is gone and may not be re- 
covered in any way. The cellulose thereafter remains 
as an amorphous body. This condition finds a parallel 
in the ultimate state of colloided nitro-cellulose, when 
the last traces of solvent are totally expelled there- 
from, and there results a pulverulent amorphous mass. 



VERY different formulae have been suggested to 
represent the composition of the nitro-products de- 
rived from celluloses, and particularly the composi- 
tion of products of maximum and minimum nitration. 
These products were, moreover, obtained by processes 
differing at the same time both as to temperature of 
reaction, concentration of acids, and the nature of the 
sulpho-nitric mixtures employed. Therefore the re- 
sults were not susceptible of any general interpreta- 

We have thought it well to take up this study 
again in a methodical manner, and to investigate the 
influence of different methods of nitration and of 
temperature upon resultant products. 


Conditions of dipping. We first decided to deter- 
mine the law in accordance with which the degree of 

*In translating this paper I have converted the chemical for- 
mulae employed therein into those of the system employed in the 
United States (O = 16). 



nitration and physical qualities vary under well-de- 
fined conditions of dipping, namely, dipping in pure 
nitric acid of various degrees of concentration and at 
a temperature of 11 C. 

The nitration was effected by immersing wadded 
cotton in from 100 to 150 times its weight of acid; 
all elevation of temperature was thus obviated and 
the strength of acidity of the bath could be regarded 
as constant during the whole of the dipping. 

The operations were conducted in large-mouthed, 
stoppered flasks of about 500 c.c. capacity and cooled 
externally by a current of water. The flasks con- 
tained about 250 c.c. of nitric acid. Three grams of 
well-carded cotton were introduced into the upper part 
of the vessel which was then corked and shaken four 
or five times, whereby the cotton was equally dis- 
tributed throughout the interior, so that each fibre 
might be considered as surrounded by 150 times its 
weight of acid. 

Percentage of nitration. The content of nitrogen 
in the nitrated product was determined by M. Schloes- 
sing's method for estimating nitrogen in nitrates. In 
order to apply this method to the determination of 
nitrogen in explosives, it had to undergo some modi- 
fications in detail, which show at the same time the de- 
gree of exactness realized in the conditions we have 
adopted. The following table presents a resum of 
the reults of our experiments : 



O w 



n to a ' 

8-5 8" 


V} ' 


O 0^ 

4J 4) C 


w .'T* 


G ^3 

rt M 


rt v 

~ .3 


2 A 

> & "Q "^ 

. G 

""* bO 

2 bo 

^ o c * J tj 

.S2 U -u 

'o "^ 




S J2 i 

"u "** 

d 1; 


t) C> *J ^ 

rt <u t) 




<U y 

E w"c3 

*SH > G 

- ^ 

w o 

Character of Specimei 

[trated product resembles c 
sly soluble in acetic ether; 
in pure ether or ether-ak 

e completely in acetic etr 
hoi. The fibre is not atta 

itrated product has the sa 
on. It becomes gelatinou 
of acetic ether and ether- 

i dissolves in the acid; pr 
id precipitable by water, 
lined swells up through th 
:r and becomes gelatinous 
Ether-alcohol produces 

roduct is extremely friabl( 
he form of a paste. Nei 
ether-alcohol produces an 

ue becomes more and me 
ngly blackened by the acti 
. Nitration, insignificant 

c ~~ 


C tj C3 

O 3 1-N W* 

O^ *"* i- 

*"rt ^ C3 


<o o.2 

ii CT rj *5 r* 

c o 

<- ^3 o 


'o rt 

r; U - 

^^ 4>.S 

r^ .^-. C 

*\ M **I* 








M <> 

Tf CO t^ 

M oo o r^ 


pi t>. 

Tf r^ co 




od N vncd 
W W i^ O 







q q 


CT < 

OO o 


K : K 


ffi = K 




co n 


l* O 

O O oo 


vO OO 

t^ M 

w ir> 

iO f"^* O 



M M 

c3 ci 


W W C^i 




+ + 


+ + 

++ + 




6, 6 


6 . 6 

66, 6 




^ - 5 

z j5 

55 - ^ 

Z 'Z " /5 




K E 


E E 








t^* r^ vo 

to o u"* o 
O O tn r> 

M O 

Tf CO 

I 1 *! 


The number of cubic centimetres indicated in the 
third column corresponds, for each degree of concen- 
tration of the acid taken, to the maximum nitration. 

Determination of the maximum nitration. This 
maximum was determined in each case by the analy- 
sis of specimens exposed to times of dipping of in- 
creasing lengths. The limit is, moreover, very clearly 
indicated by employing a solution of iodine in iodide 
of potassium, which produces a black or greenish dis- 
coloration of nitro-products containing traces of non- 
attacked cotton. We may state that, beyond the 
point where the discoloration ceases to be produced, 
prolongation of dipping does not increase the degree 
of nitration. 

Thus a specimen dipped in acid of density 1.488 
gave at the end of 24 hours of dipping 161 c.c., and 
was discolored by iodine. At the end of 70 hours it 
gave 165.7 c - c '> without discoloration. 

On the other hand, a specimen dipped in acid of 
density 1.490 ceased to be discolored by iodine after 
a dipping of 24 hours and gave 183.7 c - c - At the 
end of 128 hours the same specimen gave 183.8 c.c. 

Nature of the reactions ; speed of reactions. The 
durations of dipping which determine maximum nitra- 
tion vary considerably with the degree of concentra- 
tion of the acid. The rapid action for the density of 
1.500 (at the most from two to three hours) becomes 
gradually slower, and requires for the density of 1.483 
as much as 120 hours. The corresponding nitro-prod- 
ucts practically preserve the appearance of the original 
cotton, but for a density of about 1.470 the action is 


completely modified, the cotton swells and dissolves 
almost instantly, transforming the acid into a clear, 
transparent collodion. If this syrupy mass is poured 
into running water small white flocks are obtained, 
opaque and brittle after drying, and which preserve 
nothing of the primitive fibre of the cotton. In these 
conditions the limit of nitration is rapidly reached. 
Acid of density 1.469 gives after 5 minutes, 134.7 c - c - 1 
after 30 minutes, 140.5 c.c. ; after 20 hours, 139.3 c.c. 

When the density of the acid falls below 1.46, solu- 
tion does not occur; the action becomes much slower 
and the cotton appears unattacked, but it is shown 
upon washing that the fibres become very friable. A 
specimen was collected in the form of a paste. At 
the same time the yields decrease very considerably 
below the theoretical figure, which indicates that the 
cotton has been partially attacked. 

For densities below 1.450 it is no longer possible to 
isolate, however long the dipping, a product that is 
not blackened by iodine. The cotton is slowly trans- 
formed into products non-precipitable by water; e.g., 
at the end of 15 days there is obtained by treatment 
in water a very small residue which is intensely colored 
black or blue by iodine, and which is of a very feeble 

Discontinuities in the progress of nitration. The pre- 
ceding table shows that, under the above-mentioned 
conditions of dipping, the degree of nitration of the 
cotton increases more or less gradually, in accordance 
with the concentration of the nitric acid, from 108 c.c. 
to about 128 c.c. ; the degree of nitration then rises 



to 140 c.c., corresponding to a very small variation 
in the strength of the acid, and it remains at this fig- 
ure while the concentration of the acid increases to a 
very notable extent. It again rises under similar 
conditions to 165 c.c., then to about 180 c.c., and 
then increases gradually to the limits of nitration of 
gun-cotton properly so-called. 







The above diagram, obtained by expressing the 
density of the acid by abscissae and the corresponding 
percentage of nitration by ordinates, illustrates the 
character of the progress of the reaction. 

The existence of these discontinuities is of great 
importance in relation to the establishment of the 
chemical formulae of the nitro-derivatives of celluloses. 
It has therefore been deemed useful to reproduce 


these different degrees of nitration under entirely dif- 
erent conditions, employing sulpho-nitric mixtures. 


We employed sulpho-nitric mixtures formed of 
ordinary sulphuric acid (density 1.832) and ordinary 
nitric acid (density 1.316). The conditions of dipping 
were identical with those which have already been 
described; 3 grams of wadded cotton in 250 c.c. of 
the mixture. The temperature varied from 19 C. 
to 21 C. 

The following table presents a resum of our exper- 
iments. The first column shows the proportion by 
volume of nitric acid to sulphuric acid taken : 

Preparation of 
H.SO 4 by volume 
for i vol. of HN0 3 
A = 1.316 

No. of c.c. of NO, 
evolved from i gm. 
of the Nitrated 
Product at o C. 
and 760 m.m. , 

Character of Specimens 



184.6 ) 

Cotton not attacked. Soluble in 
acetic ether and in ether-al- 


i66. 7 V 
166.0 ) 

The cotton is very slightly at- 
tacked (a little stringy). Sol- 
uble in acetic ether and gen- 
erally becomes gelatinous by 
the action of ether-alcohol 

1. 10 

141.2 ) 
143-5 f 

Rendered gelatinous by acetic 
ether. Is only swelled up by 


133.3 (. 

Friable products 



These results give rise to various observations. 

The sharp advances in the degree of nitration indi- 
cated in the first method are equally to be observed 
under the second, and practically for the same con- 
tents of nitrogen, corresponding to a yield of 130 c.c., 
140 c.c., 165 c.c., and 180 c.c. of nitrogen dioxide, 

Thus the degree of nitration remains the same for 
proportions of sulphuric acid of 2, 1.70, and 1.50; and 
lowers abruptly by 1 8 c.c. for the proportion of 1.40. 
Nitration remains stationary for the proportions of 
1.40, 1.30, and 1. 20, and lowers again abruptly for 
the proportions and i.oo. 

In order to follow this phenomenon more closely we 
undertook dippings with proportions of acid interme- 
diary between those for which the abrupt changes had 
been observed. 

The specimens thus obtained gave the following 
results : 

Preparation of 
H-SO 4 by volume 
for i vol. of HN0 3 . 
A = 1.316. 

No. of c.c. of NO, 
evolved from i gm. 
of Nitrated Product 
at o C. and 760 

Character of Specimens 



Cotton not attacked. Soluble in 

1. 15 


Cotton attacked slightly 

The percentage of nitration is interpolated exactly 
between those indicated above. These trials confirm 
the preceding experiments, which show that the exact- 
ness of the mixtures under the conditions of our ex- 


8 9 

periments may be relied upon and that the discontin- 
uities indicated do not arise from accidental conditions. 
They show, moreover, that there are not, properly 
speaking, sharp advances in nitration ; and that there 
exists a very restricted zone of acid mixtures with 
which one may obtain intermediate nitration. But 
there exists, nevertheless, a discontinuity in the prog- 
ress of the reaction ; and the following diagram, 
obtained by expressing the percentages of sulphuric 
acid in a mixture, as abscissae, and the yield of nitro- 
gen in c.c., as ordinates, allow us to keep track of the 
progress of the reaction : 



7 8 9 JO 11 12 13 14 15 16 17 18 19 20 21 

It appears rational to admit that definite products 
correspond to the periods of constant nitration, and 
that the mixture of the two products only occurs in 
the intervals of transition. 


The properties of the cellulose nitrates obtained by 
the second process, in respect to solvents such as 
acetic ether and ether-alcohol, are, for a given degree 
of nitration, identical with those of the nitro-cellulose 
obtained by dippings in pure nitric acid. The prop- 
erties of cellulose nitrates appear then connected with 
their chemical composition, and are independent of 
the method of preparation. 

We have been able to establish this fact in a more 
rigorous manner by nitro-celluloses capable of conver- 
sion into collodions. 

The zone of collodions, according to our experi- 
ments, is narrow. It comprises celluloses the nitrogen 
content of which is comprised between 180 c.c. and 
190 c.c., approximately. 

A little above this limit however, up to 195 c.c., 
and a little below, to 166 c.c. and often to 150 c.c., 
celluloses occur which are transformed into jellies by 
the action of ether-alcohol, and which filter by press- 
ure through cloth. But it has been found that these 
substances do not produce fluid and limpid collodions 
useful in the arts; and on drying, collodions of low 
nitration have been found to yield an opaque and 
brittle film. 

Now, all the nitro-celluloses capable of producing 
collodions, obtained from widely different sources, and 
which we have had occasion to examine, are capable, 
in accordance with their composition, of being arranged 
in the method that our experience has made clear to 
us. This is shown in the following table : 


Collodion-cotton, Rousseau process, 183.2 c.c. 

" Billault-Billaudot 183.8 " 

High-temperature cotton Rousseau, 1st lot 184.0 " 

" " " " 2d " 189.0 " 
Swedish gun-cotton for the manufacture of 

gelatin-dynamite r 93-O " 

Swedish soluble cotton made at Vonges, 

in accordance with the directions in the 

aide-memoire de la marine 185.4 " 


The higher limit of the degree of nitration does not 
appear in the preceding tables, which do not present 
results relating to a density of nitric acid higher than 
1.502. Greater densities are difficult to obtain and, 
moreover, our first trials with an acid of 1.516 strength 
at a temperature of 15 C. resulted in a very marked 
attack upon the cotton. It is probable that dipping 
in acids of so great a concentration would only give 
good results at lower temperatures, but the upper limit 
of nitration of gun-cotton may be obtained indirectly 
by the use of sulpho-nitric mixtures. These mixtures, 
which were employed for the preparation of gun-cot- 
ton for military uses, produce products yielding from 
208 c.c. to 212 c.c. of nitrogen dioxide per gram. 

By proceeding with special care, we have been able 
at a temperature of 11 C. to exceed these limits and 
to obtain a product of 215 c.c. This limit remains 
exactly the same whatever the proportions of nitric or 
sulphuric acid employed, even when Nordhausen sul- 
phuric acid is substituted for the monohydrate. 

9 2 


The influence of a great excess of sulphuric acid 
affects exclusively the rapidity of the process of nitra- 
tion, which is thereby considerably diminished. 


The preceding results related to celluloses of maxi- 
mum nitration, and the iodine reaction above indicated 
permits us to regard the latter as homogeneous bodies ; 
but it is possible, by stopping the reaction arbitrarily, 
to obtain an idefinite number of celluloses of the 
same degree of nitration and of variable properties 
mixtures of different nitro-celluloses and unattacked 

Among these products we have made a special study 
of those which were furnished by the sulpho-nitric 
mixtures. On increasing the proportion of sulphuric 
acid in the mixture we may so slow down the reaction 
as to obtain for a given duration of dipping, say 
15 minutes, any degree of mean nitration desired. 
The following table shows the composition of speci- 
mens submitted during variable times to the actions 
of different acid mixtures : 

Composition of the Mixtuies 

Vol. of Nitrogen Dioxide 
evolved per gram 

HNO 3 

H 3 S0 4 

After 15 min. 

After 3 hours 

I volume 
I volume 
I volume 
I volume 

3 volumes 
3 volumes 
4 volumes 
5 volumes 




All specimens which have been exposed to only fif- 
teen minutes of dipping are blackened more or less 
strongly by iodine, which indicates the presence of 
variable quantities of non-attacked cotton ; the second 
column of the table appears to show that the excess of 
sulphuric acid only affects the speed of the reaction, 
and that all products tend toward the same limit of 

Moreover, we may verify the fact that from the very 
beginning of the reaction the products obtained are of 
maximum nitration, and that the results of rapid dip- 
ping in concentrated acids are simply mixtures of non- 
attacked cotton and products of the highest nitration. 

Thus the sample which yields 126.8 c.c. treated in 
acetic ether loses 59.44 percent, of its weight, and the 
residue presents all the characteristics of almost pure 
cotton; combustion, and coloration by iodine (black 
coloration very feeble in the presence of ferrous 
salts and hydrochloric acid). Only 60 per cent., 
then, of cotton is nitrated in this specimen ; and if we 
compare the total volume of nitrogen dioxide, 126.8 
c.c., with that of the cotton nitrated, we obtain 
214.7 c - c - 5 which is the number that we would have 
found if the reaction were allowed to proceed to com- 

This method of dipping preserves the cotton fibre 
intact, which may be of importance in relation to the 
manufacture of nitro hunting powders, for it produces 
a natural and absolutely intimate mixture of gun-cot- 
ton and ordinary cotton. Moreover, the nitrated 
material obtained in this mixture does not differ at all 


from ordinary gun-cotton, and we may hope that it 
possesses a stability equal to that of the latter. 


Composition and formula of nitro-cellulose. The 
first degree of complete nitration which the above-indi- 
cated reactions allow us to determine corresponds to 
a yield of 108 c.c. of nitrogen dioxide per gram, ap- 
proximately. This reaction is produced with pure 
nitric acid and three equivalents of water. It is slow, 
and only gives a small yield, on account of the cotton 
being attacked under the hydrating influence of the 
acid, which is then the principal phenomenon. 

As the concentration of the acids is increased, the 
per cent, of nitration appears to advance progressively 
to 128 c.c. ; at least this is what the nitrogen content 
of specimens 12 and 13 of the first series (page 83) 
would appear to show. 

It would be well, however, to maintain a certain 
reserve upon this point, because since these products 
are insoluble in the solvents, the iodine reaction re- 
mains as the only index of complete nitration ; and it 
is to be feared that certain traces of incompletely 
nitrated products lower the percentage of mean nitra- 

From 128 c.c. the percentage of nitration rises by 
marked increases to 140 c.c., then to 165 c.c., and 
finally to 180 c.c. Then, as the concentration of 
acid increases, the percentage of nitration rises pro- 
gressively, by insensible degrees, to the limit corre- 
sponding to military gun-cotton, which yields 215 c.c. 


In this last period, although there is no brisk change 
in the percentage of nitration, we may note the very 
clear transition point at 190 c.c. and 195 c.c., corre- 
sponding to the limit of the zone of the collodions. 

In order to account completely for the different 
changes by the production of nitro-products corre- 
sponding to definite formulae, it is necessary to quad- 
ruple the equivalent of cellulose. The nitro-celluloses 
which this formula leads to correspond to the theo- 
retical yields of nitrogen dioxide per gram of material 
indicated in the following table, and with which we 
compare the percentages of limiting nitration ob- 
served, corresponding to the discontinuities of which 
we have spoken already, or else to a change in physi- 
cal properties: 

Theor y ment 

C 24 H 29 O 20 ^NO 2 )i 1 . .Cellulose endecanitrate ) _ j 214 c.c. 215 c c. 

C 24 H 30 O 20 (NO 2 ), . Cellulose decanitrate.. ' 
C 24 H 31 O 20 (NO 2 ) 9 . . .Cellulose enneanitrate 
C 24 H 32 O 20 (NO 2 ) 8 .. .Cellulose octonitrate. . 
C 24 H 33 O 20 (NO a ) 7 ... Cellulose heptanitrate. 
C 24 H 34 O 20 (NO,) 6 . .-Cellulose hexanitrate. 

C 24 H 36 O 20 (NO 2 ) 6 . ..Cellulose pentanitrate 

I Collodions. ... ) 

203 215 

V Friable cottons < 146 
) I 128 



C 24 H 39 O 20 (NO 2 ) 4 . . . Cellulose tetranitrate 108 

It will be seen that these formulae take suitable ac- 
count of the production of limits of nitration and of 
all particulars presented by the reaction. 

The approximate results obtained, however, are 
always lower than the exact volumes of nitrogen by 
about Y^J-. The differences appear to be attributable 
to the presence of a very small quantity of products 
of lower nitration. 

Properties of nitro-celluloses. The explosive prop- 
erties of nitro-celluloses are in direct relation to the 


percentages of nitration. As the nitration diminishes 
the vivacity of combustion in open air decreases, and 
the production of carbonaceous residue becomes more 
marked. We may thus classify at a glance nitrated 
celluloses in one of three groups gun-cottons, collo- 
dions, or friable cottons. The measurements of pres- 
sures developed by these different products in a closed 
chamber show that the force similarly diminishes with 
the percentage of nitration. Thus a collodion-cotton 
yielding 184 c.c. produces pressures inferior by -J- to the 
pressure furnished by Moulin-Blanc gun-cotton afford- 
ing 21 1 c.c. The percentage of nitrogen constitutes 
a true measure of the explosive qualities of a product. 

Finally, we may mention that the stability of nitro- 
celluloses decreases with the percentage of nitration, 
with respect to reagents such as hydrochloric acid 
and ferrous salts. For products of low nitration the 
reaction commences when cold ; for those of mean 
nitration a few moments' heating is required, but for 
celluloses yielding more than 200 c.c. of nitrogen 
dioxide per gram, the attack commences only after 
sustained ebullition. These products appear then to 
acquire the maximum of stability along with the maxi- 
mum of power. 

PARIS, September, 1883. 



By Professor D. MENDELEEF 

(Translated from the Russian by Lieutenant John B. Bernadou 
U. S. Navy) 

THE very favorable results obtained with pyro- 
collodion, and its adaptability to arms of all calibres, 
depend upon its composition and properties, which, 
for purposes of illustration, may be compared with 
those of other materials employed as smokeless pow- 

As to its chemical composition, pyrocollodion may 
be designated homogeneous,* and herein consists one of 
its most important qualities. All previous and pres- 
ent forms of powder did not have or do not have this 
property to the degree here implied. From their 
very method of preparation, black and brown powders 

* Homogeneity, in its full chemical significance, is not claimed, 
inasmuch as the composition of cellulose itself remains a matter 
of doubt. The quality is urged from the technical standpoint, in 
relation to the properties of other smokeless powders. It is pos- 
sible that a solvent may be found capable of separating pyro- 
collodion fractionally; but pyrocollodion insoluble in ether or 
alcohol, but soluble in a mixture of these substances, is far more 
homogeneous than other forms of nitro-cellulose or any of the 
nitro-glycerin powders, inasmuch as the latter are readily capable 
of fractional subdivision. 



are coarse mechanical mixtures, for which any consid- 
eration of homogeneity is out of the question. The 
same is true for those smokeless powders containing 
ammonium nitrate, picrates, etc. Nitro-glycerin 
powders may be regarded as gelatinous solutions of 
nitro-cellulose in nitro-glycerin, which, from their 
composition, are, chemically, non-homogeneous; more- 
over, various solvents (alcohol, ether, acetone, etc.) 
dissolve certain constituents out of them, leaving 

The same may be said of present-day types of nitro- 
cellulose powders ; alcohol dissolves out of them the 
nitro-celluloses of lower nitration ; a mixture of ether 
and alcohol, the collodions, leaving the excess of 
highly nitrated cellulose undissolved. Pyrocollo- 
dion,* however, surrenders no part of its substance to 
alcohol, while it is wholly soluble in a mixture of 
ether and alcohol. This chemical homogeneity of 
pyrocollodion, taken in the sense in which it is stated 
to be employed, plays an important role in its com- 
bustion ; for there are many reasons for believing that 
in the case of the combustion of those physically but 
not chemically homogeneous substances, such as the 
nitro-glycerin powders (ballistite, cordite, etc.), the 
nitro-glycerin portion is decomposed first, and the 
nitro-cellulose portion burns subsequently, in a differ- 
ent layer of the powder:f It is to be added that the 

* Under the assumption that the remainder of the solvent is 
wholly expelled from the powder. 

f The experiments of Messrs. I. M. and P. M. Tcheltsov at the 
Scientific and Technical Laboratory show that for a given density 


homogeneity of pyrocollodion possesses a direct bear- 
ing upon the uniformity of ballistic results developed 
by its use. 

Besides chemical homogeneity, pyrocellulose and 
the powders prepared therefrom possess a second dis- 
tinguishing quality, viz., that for a given weight of 
their substance they develop a maximum volume of 
evolved gases, the latter being measured at a given 
temperature and pressure. This new conception in- 
volves certain intricacies and complexities, and may be 
discussed to some degree of fulness, in what relates to 
nature and volume of gases evolved upon the decom- 
position of powder. 

According to the law of Avogadro-Grard, * the 
chemical equivalents or quantities of matter expressed 
by simultaneous chemical formulae (e.g., H a O = 18, 
water; CO = 28, carbonic oxide; CO 2 = 44, carbonic 
acid; N 2 = 28, nitrogen) occupy at a given temper- 
ture and pressure a volume equal to that occupied by 
two parts by weight of hydrogen, H a = 2 (its molecu- 
lar equivalent). Consequently, if we possess the full 
chemical equation of combustion of a substance or 

of loading, the composition of the gases evolved by nitro-glycerin 
powders varies according to the surface area of the grains (i.e., 
the thickness of strips or cords), a phenomenon not to be ob- 
served in the combustion of pyrocollodion powder. There is 
only one explanation for this, viz., that the nitro-glycerin, which 
possesses the higher rate of combustion (Berthelot), is decom- 
posed sooner than the nitro-cellulose dissolved in it. This is the 
reason why the nitro-glycerin powders destroy the inner surfaces 
of gun-chambers with such rapidity. 

* The development of this law is given in " Principles of Chem- 
istry," by D. Mendeleef, 6th ed., 1895, chap. 7. 


mixture of substances, of which the products are 
gases or vapor, it is easy to calculate the volume 
occupied by these products at a given temperature 
and pressure. For example : the combustion of black 
powder may be expressed typically by the equation : 

2KNO.+ S + 3 C =K 3 S + 3 CO a + N 2 . 
Mol. wt, 2 X 101 + 32 -{-3 X 12 = 110 + 3X44+28=270. 
Volume in form of gases, 3X2 + 2 = 8. 

That is, for 270 parts by weight of powder ingredients 
8 volumes of gas* are formed, or 29.6 volumes per 

* If the weights of the equivalents be expressed in grams we 
may ascertain the volume of gas evolved in litres, when the pres- 
sure P (in millimeters of the mercurial column) and the temper- 
ature / (in degrees Celsius) are known. Thus as two equivalent 
weights of hydrogen and the equivalent of each gas occupy at a 
temperature t = o and a pressure P = 760 mm. a volume of 2? 
litres, then for / and P this volume becomes 

22* (i + 0.00367 /) 25. 

Consequently one volume, expressed in grams, occupies, approx- 

11.1+0.4070* litreSf 

where/ corresponds to the number of atmospheres, each of 760 

mm. atoC. ; i.e., / = -;. Thus, in our example, if t = 2000 C. and 

p = 2500 atm., the 8 volumes of gas produced by the combustion 
of 270 grams of powder occupy an actual volume of 

g< 11.1+0.0407.2000 = fi Htre> 


At o C. and a pressure of i atmosphere we attain a volume of 88.8 
litres for 270 grams. In this manner it is easy to proceed to the 
value ^1000 given in the text, the actual volume of evolved gases 
per kilogram of powder. 


1000 parts by weight of explosive. Brown powder 
(cocoa) represents (in its greater progressiveness of 
combustion and in certain other respects) a partial 
transformation from black to smokeless powders, and 
is characterized by the partial carbonization of its 
charcoal (which contains much hydrogen and approx- 
imates to a composition C 6 H 4 O), and by the small 
quantity of sulphur entering into its composition. Its 
mode of combustion may be expressed approximately 
as follows: 

6KN0 1 +2C 6 H 4 0= 3 K S CO,+7CO 

6X 101 + 2 X 80 = 3X138 + 7X28+4X18+3X28=766. 
Volume of gases, 7X2 +4X2 +3X2 =28. 

^, = 36.5- 

Or as follows : 

4 KN0 3 +C 5 H 4 0+S=K 2 S0 4 +K 2 C0 3 + 4 CO+2H 2 0+2N 2 , 
Mol. wt. = 516, Volume of gases = 16, 

^,.., = 31. 

It is evident, then, that the gas volume correspond- 
ing to brown powder is nearly 34, greater than that 
for black powder; whence originates the preference 
generally accorded to brown powder over black. 

In a similar manner we have for the complete 
(typical) combustion of nitro-glycerin, 

4 C 3 H 6 N 3 9 = i2C0 3 + ioH 3 + 6N 3 + O,. 
Mol. wt, 4 X 227 = 12 X 44 + 10 X 18 + 6X28+32 = 908 
Volume of gases = 12X2 +10X2+6X2+2=58. 

^ooo = 63.9- 

If we present in the same manner the decomposi- 
tion of a type of nitro-cellulose of high nitration, such 


as Abel's trinitro-cellulose, C 6 H 7 (NO 3 ) 3 O 6 , we obtain 
the following: 

2C.H,N 1 1I = sCO, + 9 CO + 7 H 2 0-f 3 N 2 . 
Mol. wt., 2 X 297 = 3 X 44+ 9 X 28 + 7 X 18+3 X 28 594. 
Volume of gases = 3X2 +9X2 -{-7 X 2 +3 X 2 =44. 

^,000 = 74-1- 

If the typical combustion of this nitro-cellulose be 
presented by the equation 

2C 6 H t N.O n = ioCO a + 2CO + 7 H, + 3 N 3 , 

that is, if we assume that water is not formed, and 
that the oxygen combines wholly with the carbon, 
then the F, 000 remains unchanged, as the volume of 
gas formed (22X2= 44) remains as before. Therefore 
we need not stop to consider how the oxygen is dis- 
tributed between the carbon and hydrogen in the 
products of combustion, as the value of F" 1000 does not 

If, for nitro-cellulose of high nitration we substitute 
Eder's pentanitro-cellulose, C 12 H 16 (NO a ) 6 O 10 , which 

* It is another matter if a portion of the oxygen continues in 
combination with the nitrogen, or if the oxygen proves insuffi- 
cient to convert all the carbon and hydrogen into gases; that is, 
if hydrocarbons are formed; but this becomes a case of incom- 
plete combustion. Such conditions have a certain bearing upon 
the combustion of smokeless powders, especially when the sol- 
vent is not completely expelled; but, on the one hand, the quan- 
tity of hydrocarbons formed is relatively small, and on the other, 
they are formed (as also compounds of carbon and nitrogen, as 
cyanogen), in relatively small quantities, for all powders, even 
when the latter contain an excess of oxygen. In considering 
type forms of combustion there is no need of investigating second- 
ary conditions of this class, especially as by so doing we are 
diverted from the direct study of the general problem. 


corresponds to a content of 12.75 P er cent, nitrogen, 
and to the composition of ordinary nitro-cellulose 
employed for smokeless powders, we obtain a greater 
evolution of gas, for 

2C 12 H lb N 6 0, - CO, + 2 3 CO + i 5 H,0 + 5 N.. 
Mol.wt., 2X549- 44+ 23 X 28 .+ 15 X 18+5X28=1098. 
Volume of gases = 2 + 23.2 -j- 15.2 +5.2 = 88. 

r n = 80. i. ' 

The increase in volume of gas hereby realized is due to 
the fact that the quantity of carbonic acid evolved is 
diminished, while that of carbonic oxide is increased, 
which causes an increase of total gas volume, since, 
for the equation 

C + 0. = CO,, r iM .=45-55 

and for the equation 

C + =CO, F 1000 = 7i.4. 

If we descend to a lower nitration and consider 
Eder's tetranitro-cellulose, C 12 H 18 (NO,) 4 O, , we have 
no longer the case of complete combustion; for 20 
equivalents of oxygen are required to convert 12 of 
carbon and 16 of hydrogen into gaseous products and 
vapors, while there are but 18 of oxygen available, un- 
less we assume the products of combustion as CO, 
H,O, and H 9 .* It is known, however, that in the case 

*In the latter case (without formation of carbon) the decompo- 
sition would be: 

C )a H 16 N 4 18 = I2CO + 6H 3 0+ 2 H a + 2N a . 
Mol. wt., 504 = 12 X 28 + 6 X 18 + 2 X 2 + 2 Xj28 = 504. 
Volume of gases = 12X2 +6X2 +2X2 + 2X2 =44. = 87.5. 

But typical combustion, according to such an equation, is prac- 
tically impossible; carbon and hydrocarbons are formed, and the 

| UNiVR ? 


of combustion of carbohydrates low in oxygen, the 
latter combines with the hydrogen, from its greater 
affinity for that substance, leaving a part of the carbon 
deposited in an uncombined state. Consequently, such 
conditions do not correspond to complete combustion. 
The formation of CO, shows that there is a certain 
excess of oxygen in pentanitro-cellulose ; whereas 
typical combustion, corresponding to maximum gas 
volume, requires all carbon to be converted into car- 
bonic oxide. Such typical combustion is afforded by 
pyrocollodion, the composition of which corresponds 
to the formula C 80 H 98 N ja O 49 , as is shown by the follow- 
ing equation : 

5C.H..O. + I2HNO, = C,.H,,(NO.) 11 1 . + "H.O. 

Cellulose Nitric Acid Pyrocellulose Water 

combustion that actually does occur is intermediate in character 
to that expressed by the above and by the two following equa- 
tions : 

2C ia H 18 N 4 O, 8 = C 4 -f- 2oCO -f i6H a O + 4N a . 
2C ia H 19 N 4 18 = 22CO -f i 4 H a O + C a H 4 + 4N a . 
For the first, Kiooo = 79-45 for the second, 81.4; the mean of 
the two latter is 80.4; of all three, 82.4. This quantity is close to 
that afforded by pentanitro-cellulose and pyrocollodion. In this 
manner may be explained the phenomenon that upon the combus- 
tion of nitro-cellulose containing a little less than 12.5 per cent, 
nitrogen, or of pyrocollodion containing a certain quantity of un- 
evaporated solvent (which is equivalent to a lowering of nitra- 
tion), results are obtained that approximate to those produced by 
pyrocollodion powder, although velocities and pressures are 
somewhat lowered. The existence of this phenomenon depends 
upon the homogeneity of the powder; whence it follows that it is 
better to have a content of a little less than 12.5 per cent, nitrogen 
with a homogeneous powder than a content of above 12.5 per cent, 
with the powder non-homogeneous, and that the best results are 
developed by homogeneous pyrocollodion of nitration N = 12.4 
per cent. 


In typical combustion it corresponds to the follow- 
ing equation : 

C SO H 38 N 12 48 = 3 oCO + i 9 H 2 0+6N r 

Mol. wt., 1350 30X28+19 Xi8 + 6 X 28 = 1350. 

Volume of gases =30X2 +19 X2+6X2 = no. 

Before proceeding further, we desire to call atten- 
tion to the fact that, whereas for brown powder we 
realize a volume of 34 approximately, we have here a 
volume of 81.5, whence, judging from volumes of 
evolved gases, pyrocollodion should prove 2j times 
more powerful than brown powder. Actual experi- 
ments show that the powders stand in about 'this rela- 
tion to one another. In units of energy per unit for 
weight of explosive we have 

Pyrocollodion Brown Ratio 

powder powder 

47 m.m. R. F. about 220 81.5 2.7:1 

9-in. gun, " 223 90.7 2.5 : I 

12-in. gun, " 210 93 2.3:1 

From this it is evident that our computed value of 
F" loou is in complete accordance with actual experimental 

In this manner we may consider it as proven that, 
for a given temperature and pressure, pyrocollodion 
develops a greater volume of gases (and vapor) than is 
developed by black or brown powder (for which 
J^iooo = 3)> an d even greater than is afforded by pow- 
ders prepared from the more Jiighly nitrated forms of 
nitro-cellulose, and by the nitro-glycerin poivders* 

* The rapid, simple and novel method of comparing the force of 
explosives herein employed was first suggested and used by me 


In order to establish the full significance of the 
above deduction, it remains to show (i) that from the 
standpoint of practical applicability we can foresee no 
other material capable of developing as great a value 
for F 1000 as pyrocollodion ; (2) that the physico-chem- 
ical and ballistic qualities necessary in a smokeless 
powder are developed by pyrocollodion, not in a less, 
but in an equal or greater degree than by other mate- 
rials employed as smokeless powders; (3) that the 
estimate of ballistic efficiency of a smokeless explosive, 
through consideration of its volume of evolved gas, 
without regard to conditions of temperature, leads us 

in 1892. It was developed through comparison of the composition 
of smokeless powders and of their products of combustion and of 
the results of experimental firings made at the laboratory. Al- 
though I see clearly that not only the volumes of products of 
combustion, but also their temperatures, must be taken into 
account in a complete analysis of phenomena attending the 
decomposition of smokeless powders, nevertheless I purposely 
give preference in these investigations to phenomena relating to 
volumes of gases evolved, not only on account of the simplicity 
of the latter and their direct accordance with ballistic results, but 
for the reason that with present methods for estimating tempera- 
tures developed by explosives (and these methods are unreliable) 
it becomes necessary to make numerous arbitrary assumptions 
(especially in relation to specific heats of gases and vapors at high 
temperatures); while for volume calculations we have definiteness 
of composition as a starting point; and if there be any assump- 
tion to be made, it relates to the distribution of the oxygen be- 
tween the carbon and hydrogen, which, from the chemical stand- 
point, is not so material and so far as it relates to volume it 
is of little significance. In all cases, however, I present but the 
elementary comparisons of performances of powders, as the fuller 
treatment of the subject does not constitute the object of my 


into no error, although it would at first appear that 
temperature would have a direct effect upon the prac- 
tical qualities of a powder. 

Since smokeless powder was discovered, so many 
schemes were set afloat for meeting general demands, 
that up to the present time there remain as open 
questions which form of smokeless powder is the 
superior, and whether new and still more efficient forms 
may not be looked for in the future. In order to re- 
ply to these inquiries, it will be necessary first to glance 
over the compositions of materials capable of conver- 
sion into smokeless powders under the following as- 
sumptions: (i) That they leave no solid residue after 
combustion, and that their gases exercise no injurious 
effect upon the metal of guns ; (2) that they undergo 
no change upon keeping for long periods of time, and 
contain no volatile ingredients ; and (3) that they may 
be readily prepared in quantities sufficiently abundant 
for practical needs. 

There are but few elements capable of producing 
gases that do not act upon metals, and, generally 
speaking, it is useless to try to find others besides 
hydrogen and nitrogen, and their compounds with 
oxygen and carbon, that do not act upon gun-cham- 
bers at the temperatures of combustion of powder. 
Therefore, in general terms, the composition of those 
mixtures or compounds suitable for conversion into 
powder may be expressed as 

The energy imparted to the projectile is derived from 


the conversion of the mass of the powder into gases, * 
the transformation being accompanied with the pro- 
duction of great heat. These fundamental conditions 
serve to limit the number of materials that are capable 
of conversion into smokeless powder, the limitations 
arising not only from the above-named practical re- 
quirements, but also from the chemical impossibility of 
existence of many bodies which, if obtainable, would 
decompose in the manner requisite for efficient ballis- 
tic action. Thus, e.g., there does not exist, nor can 
we look forward to the existence of, such a polymer 
(in the solid or liquid form) of hydrogen as H , which 
would decompose into hydrogen, H a , with a corre- 
sponding production of heat.f 

If we may not look for explosives among the sim- 
plest chemical combinations of the elements, we may 
perhaps find them among those compounds of nitro- 
gen and hydrogen which stand in the same relation to 
ammonia, NH 3 , as the hydrocarbons do to methane, 

* And is not derived from an external source of energy, such as 
the tension of a spring or the physical compression of gases, as 
in the Giffard gun. 

f If such a substance existed, then its decomposition, according 
to the equation H a = wH a , would afford a weight 2 for a volume 
2; that is, Fiooo would be equal to 1000, the greatest possible 
value. For nitrogen under similar conditions we would have 
Fiooo = 71.4, or less than for pyrocollodion. But the existence 
of such a polymer is highly improbable. If argon (vid. Mende- 
leef, " Principles of Chemistry," 6th ed., 1895, p. 749) were the poly- 
mer of nitrogen, N 3 , its conversion into nitrogen could only be 
accomplished through the absorption of heat; i.e., it would find no 
place in the category of " explosive " bodies (to which ozone pos- 
sesses a relation). 


CH 4 . We should thus have, corresponding to ammo- 
nia, the series N M H w + 2 (e.g., diamide N 2 H 4 ; triamide 
N 3 H 6 ; etc.) and the series N M H M , N M H w _ a , etc. As 
the representative of the latter we have, for n = 3, the 
nitro-hydric acid of Curtius, N 3 H, which actually is a 
very explosive body, and which forms salts, e.g., with 
ammonium N 3 (NH 4 )= N 4 H 4 , which is also explosive, 
decomposing into the gases nitrogen and hydrogen 
with the evolution of heat, although ammonia itself is 
not susceptible of explosive decomposition, but ab- 
sorbs heat in the reaction. If such compounds could 
be easily prepared, and if they possessed the qualities 
necessary to an efficient smokeless powder, such as 
non-volatility, good keeping quality, progressiveness 
in combustion, etc., they would prove especially suit- 
able for conversion into smokeless powders, as the 
corresponding values of F" 1000 would be greater than 
for other powders. Thus we should have for nitro- 
hydric acid, 

2N 3 H = H 3 + 3 N,. 
Mol, wt. 2X43= 2+3x28=86. 
Volume of gases, 2 + 3X 2 = 8. 
PI..O = 93-0. 

For N W H M the volume would be still greater, as 
F 1000 = 133.3. But even if such products could be con- 
veniently prepared from readily procurable materials it 
would be 'useless to consider them as available for 
conversion into smokeless powders, for the reason 
that they do not decompose through gradual or 
progressive combustion, which is indispensable in a 


smokeless powder, but detonate or decompose with 
extreme suddenness; hence, while they might prove 
suitable for filling mines or shells, they are unadapted 
for use in cannon. This property of progressive com- 
bustion or decomposition in successive strata is pos- 
sessed only by those substances containing both com- 
bustible ingredients, and ingredients capable of effect- 
ing progressive combustion, such as carbon and 
hydrogen, which are consumed by the oxygen held in 
close proximity to them but which is not in direct 
combination with them. 

The "combustion" of a powder is the union of 
the carbon and hydrogen of the mixture or com- 
pound with the oxygen that it contains, and with 
which it is in association, but not in direct combina- 
tion. From what has been said already, it is evident 
that if the powder is to be smokeless and produce the 
maximum volume of gas, F 1000 , it must evolve no 
other gases than carbonic oxide, CO, water vapor, 
H 9 O, and nitrogen, N a . If hydrogen be evolved, with- 
out the formation of the corresponding quantity 
(equal volume) of carbonic acid, free carbon may re- 
sult, i.e., the powder will not be wholly smokeless on 
account of insufficiency of oxygen. If the combus- 
tion, as indicated by the equation, reveals carbonic 
acid or free oxygen (without the corresponding vol- 
ume of hydrogen), an excess of oxygen is evident, 
and F^ooo will not possess its maximum value. 

We have, therefore, in the case of a composition or 
mixture C M H 2w N, ? O r , the maximum volume F 1000 for 
typical smokeless combustion, corresponding to two 


conditions: (i) When the content of oxygen, r, is 
just sufficient to convert the carbon into CO, and 
the hydrogen into H a O, i.e., when r = n -f- m ; (2) 
when the content of hydrogen is relatively great, as 
F 1000 for H a O equals iii.i, i.e., more than for nitro- 
gen and for carbonic oxide, for which F 1000 = ^f f- 
= 71.4. Moreover, all substances of the composition 
C H H tm N q O H+m will develop volumes between 71.4 and 
iii.i, if the decomposition products be CO, N, and 
H a O alone, as is required for rendering F, 000 a maxi- 
mum. Our problem becomes, therefore, the com- 
parative examination of those bodies rich in hydrogen, 
for which F" 1000 may be greater than for pyrocellulose 
(81.5). We must ask: Are there not known sub- 
stances, or mixtures of substances, rich in hydrogen 
suitable for smokeless powder? To answer this query, 
let us examine various definite compounds and mix- 

Among the carbon compounds a large content of 
hydrogen is characteristic of methane (marsh gas), 
CH 4 ; also among the nitrogen compounds, or the 
ammonium derivatives. 

Hydrocarbons of the limiting (saturated) series 
C w H aw _|_ 9 * form nitro-compounds, and may, therefore, 
produce explosives. To methane itself, CH 4 , there 
correspond mononitro-methane, CH,(NO a ); dinitro- 
methane, CH a (NO a ) a ; trinitro-methane or nitroform, 
CH(NO a ),, and tetranitro-methane C(NO a ) 4 . These 

* See "The Principles of Chemistry," by D. Mendeleef, 1891, 
Longmans, Green & Co., London, Vol. I, p. 344. J. B. B. 


substances are volatile as well as explosive, but all 
represent a deficiency or an excess of oxygen. As 
shown by V. Meyer and Professor Zalinski, the ex- 
plosive properties of mononitro-methane are espe- 
cially great when it is combined with potassium or 
sodium to form the metallic salts, CH a KNO a and 
CH 5 NaNO a , which represent, so to speak, first homo- 
logues of the salts of nitric acid, since CH a NaNO a 
NaNO a equals the homologous difference CH a . 
Experiment shows that this substance belongs to 
the category of detonating explosives, and is, there- 
fore, unsuitable for use in guns (but suitable for 

If the decomposition proceeds without formation of 
free carbon (although there be but little oxygen), it 
should be as follows: 

2CH 8 NO a = 2CO + 2H a O + H a + N a . 

If it be thus, then F 1000 =98.3, which is very great. 
But, as has been said, the substance is unfit for use in 
guns on account of its tendency to detonate. Be- 
sides, like other nitro-methanes, it is volatile, and for 
this reason is further unadapted. 

The little known, but doubtlessly explosive, dinitro- 
methane contains an evident excess of oxygen, devel- 
oping on combustion, CO, -f- H a O + iO a + N a , which 
corresponds to the relatively small volume F 1000 = 66. 
It is evident that the excess of nitrogen and of oxygen 
combined with it in the NO a , according to the known 
principle, does not increase but rather diminishes 
F, 009 . The same is true for nitroform, CH(NO,) 3 = 


CHN 3 O 8 ,* and for tetranitro-methane, the discovery 
of which is due to the skill of our eminent savant, 
L. N. Shishkov. Both* of these substances contain too 
much oxygen to develop maximum gas volumes. A 
large value of F" looo would be characteristic of mixtures 
of products of nitration and of hydration (substituting 
the water radicle for hydrogen, H+OH) derived 
from methane, as : 

4CH.NO, + CH,N,0 4 - 5 CO + ;H 3 O + 3 N,f 
Mol. wt., 4X61 + io6=5X28+7X 18+3x28 = 350. 
Volume of gases, 5X2+7X2+3X2 =30. 

F 1000 = 8 5 . 

Nitro-methane Methyl alcohol 

4CH,NO,+CH 4 = 5CO +8H,O+2N,. 
Mol. wt., 4 X 77+32 = 5x28+8 X 18+2x28 = 340. 
Volume of gases =5X2 +8 X 2 +2 X 2 =30. 

F 1000 = 88.2, etc. 

But such mixtures, although possible from the chem- 
ical standpoint, are unsuitable for use as powder, be- 
cause their constituents are in part volatile ; and this, 
apart from the consideration that liquid explosives are 
prone to detonation, which is more to be dreaded than 
formation of smoke, as detonation destroys the guns. 

* As the typical decomposition of nitroform, we have 

4CHN 8 0,=4CO a + 70, -f-2H a O + 6N a . 
Mol. wt., 4 X 151 =4 X 44 + 7X32 + 2X18+6X28 = 604. 

Volume of gases 4 X 2 +7X2 +2X2 +6X2 =38. 
f The mixture yCH 4 -+- 3C(NO 2 ) 4 presents such a composition, 
etc., but such mixtures are all as practically unsuitable for pow- 
der as mixtures of mono- and dinitro-methane. 


Among the closely allied derivatives of methane as 
a hydrocarbon rich in hydrogen, the development of 
a large gas volume may be looked for from substances 
presenting the composition CN 3 H 6 O 4 . Such a com- 
position is possessed, for example, by the mixture 
of a molecule of formic aldehyde, CH a O (or of one 
of its numerous polymers), with ammonium nitrate, 
NH 4 NO 3 , or the hydroxyl derivative of methylamine 
(i.e., CH,NH a in the form CH a [OH]NH a ) in com- 
bination with nitric acid, HNO 3 . The typical de- 
composition of such a compound, if realized, would be 
expressed by the equation : 

CN 3 H 6 4 = CO+ 3 H a O +N 3 . 
Mol. wt., 1 10 = 28 +3 X 18 + 28 = no. 

Volume of gases 2 +3X2 + 2 10. 
Fioo. = 9.09. 

But such a compound either cannot be produced, or 
else is obtained only with great difficulty ; or, as a mix- 
ture of ammonium nitrate with the polymers of formic 
aldehyde (e.g., glucose, C 6 H la O 6 = 6CH a O), it develops 
undesirable qualities, such as hygroscopicity, a charac- 
teristic of all mixtures containing ammonium nitrate, 
and is therefore unsuited for use as smokeless powder. 

Hence, after searching through all the possible com- 
binations of the simplest derivatives of methane, we 
are unable to find among them (as also among sub- 
stances containing no carbon) any suitable for employ- 
ment in practice as smokeless powder, although we 
find compounds developing larger volumes of gas than 
pyrocollodion, and which may prove suitable for use 
in mines. 


If we turn from substances containing one atom of 
carbon to those with two, three, etc., atoms to the 
molecule, we shall find, other conditions being the 
same, smaller values of F" 1000 , the volume decreasing 
the farther the limit is departed from, as is illustrated 
in the following table of possible, little volatile, com- 
pound ethers of nitric acid* and their hypothetical 
nitro-compounds, corresponding to the series of alco- 
hols, C w H aw (NO a ), 

3 C 3 H 4 (N0 3 ) a + 2C.H.O, F 1000 = AV 1000=86.2 

C 3 H 6 (NO,) a " - tto 1000=84.3 

QH 8 (N0 3 ) a + 4 C 4 H 7 (N0 2 )(N0 3 ) a " = T 1000=83.7 
C 6 H 10 (N0 3 ) a + 4 C 6 H 8 (NO a ) a (NO,) a ' = T Wo 1000=82.7 

etc., etc. 

The possible, yet up to the present, hypothetical, 
nitric ethers of nitro-glucol, although capable of devel- 
oping large values of V IOOQ , and adaptable on account 
of their non-volatility, possess no advantages over 
derivatives of the higher alcohols, such as glycerin 
and mannite, materials that are readily obtainable 
as they are widely disseminated throughout nature. 
We shall therefore fix our attention upon the latter, 
first, as they present in their analogues substances ex- 
tremely rich in hydrogen, capable of producing large 
values of F 1000 ; and second, because they are easily re- 

* Considered by themselves these ethers of diatomic limiting 
(saturated) alcohols C W H 2 (NO 3 ) 2 consume into CO and H a O only 
for n = 3. For greater values of n there is a deficiency in oxy- 
gen; for n = 2, an excess. We have chosen them as an example 
on account of their slight volatility, and because they approximate 
nitro-glycerin and nitro-mannite in composition. 


acted upon by nitric acid, forming the explosive com- 
pound ethers, nitro-glycerin, C 8 H 5 (NO S ) 3 C e H 6 N 8 O fl , 
and nitro-mannite, C 6 H 8 (NO 3 ) 6 O e =: C 6 H 8 N 6 O 18 . Both of 
these nitro-derivatives are easily prepared. The for- 
mer was first employed as an explosive by the re- 
nowned Russian chemist N. N. Zinin, at the time of 
the Crimean war, and subsequently by V. F. Petru- 
shevski, in the sixties, before the discovery and very 
general employment of Nobel's dynamite and other 
nitro-glycerin preparations; the cause of their gen- 
eral use being the ease with which the base material 
glycerin was obtainable in nature, while the reac- 
tion with nitric acid (admixed with sulphuric) was 
easily effected, i.e., the manufacturing process was a 
simple one. 

Nitro-mannite, isolated and investigated by N. N. 
Sokolov, professor at the Medico-Chirurgical Academy, 
is also easily prepared, but not in its lower degrees of 
nitration. This circumstance is important, for the 
reason that the readily manufactured nitro-glycerin 
and nitro-mannite are not themselves available for use 
in guns, although very well adapted for detonating 
effects. They correspond, moreover, to relatively 
small values of F 1000 , as they contain an excess of 
oxygen : 

C s H t (NO,), affords F 10M = 63.9. 
C,H.(NO,). " F 10M = 6i. 9 . 

But as these substances contain an excess of oxygen, 
they may be admixed with others containing a defi- 
ciency thereof, and which they consume, evolving car- 


bonic oxide and developing relatively large volumes, 
Fj 000 ; while by admixture with such substances, low in 
oxygen, or not containing it, their detonating qualities 
may be caused to diminish, or made to vanish, as in 
dynamite, by combination with an inert base (tripoli, 
magnesia, etc.), whereby the tendency of nitre-glyc- 
erin to detonation through shock is diminished. In 
this manner, by admixture with a combustible sub- 
stance, nitro-glycerin powders are formed. If we take 
cordite as an example, we find that on account of 
its excess of oxygen it produces a relatively small gas 
volume ; we may, therefore, select a mixture of nitro- 
glycerin and collodion (assumed as C e H 8 [NO a ] a O 6 ) 
such as ballistite and determine for it F" 1000 on the as- 
sumption that it shall develop only CO, H a O and N t . 

2C,H 6 N 3 9 +7C B H 8 N,0 9 =48CO 

Mol.wt. 2X227+7 X 252=48X28+33X18+10X28 = 2218. 
Volume of gases =48X2 +33X2 +10X2 =182. 

Therefore, if nitro-glycerin powders contain only 
the quantity of nitro-glycerin necessary to produce 
H a O and CO, then the volume of gases evolved by 
them is almost the same as that developed by pyro- 
collodion. It is evident, then, that neither nitro- 
glycerin nor its mixtures, when employed as smoke- 
less powder, evolve volumes of gases greater than pyro- 
collodion, and that admixture with other substances, 
of whatever kind they may be, although homogeneous 
from the mechanical standpoint, are still far less homo- 
geneous than any single substance, and that it is use- 


less to seek for nitro-glycerin powders capable of ex- 
ceeding pyrocollodion powders in point of magnitude 
of f 1000 , apart from other considerations. This ap- 
plies also to nitro-mannite, the source of preparation 
of which is far less common than glycerin, and to 
many of the hydrocarbons analogous thereto, as glu- 
cose, starch, cellulose, etc. If all of the six atoms of 
carbon in mannite are in the same combination as in 
the limiting (saturated) alcohols : 


then in glucose, C 6 H 13 O 6 , one atom of carbon should 
be combined as in aldehyde, 


and, therefore, if mannite represents a hexanitrated 
product (compound ether as derived from an alco- 
hol), glucose represents only a pentanitrate. Mate- 
rials such as cellulose, starch, and the like, of compo- 
sition C 8 H 10 O 6 , may be regarded as the preceding alco- 
hols anhydrous, arranged as follows with reference 
to the di-alcoholic groups: 


C.H 10 6 = COH CH(OH) CH(OH) CH(OH) , , 


whereby it appears that there are only three com- 
plete alcohol groups remaining out of the six in man- 
nite. Therefore, in the latter case, we should expect 
to find only trinitrated products, which is actually 
what occurs, If such a scheme throws light on the 


matter from one standpoint (as relates to the number 
of hydroxyl radicles giving rise to nitric ethers), it illu- 
minates it obliquely from another, which is of consid- 
erable importance to us. In all aldehydes, beginning 
with the formic and acetic, a tendency to polymeriza- 
tion is to be noted, due, doubtless, to the property 
of aldehydes of entering into various combinations 
(with H 3 , O, NaHSO 9 , etc.); hence the composi- 
tion C 6 H ln O 5 , containing an aldehyde grouping, 
should also possess this property, so far as relates 
thereto. We may, therefore, safely assume that the 
molecular composition of cellulose, judging from its 
properties, is polymerized, i.e., it is of the form 
C eM H, ow O 6M , where n is probably very great. If we as- 
sume n = 5, the cellulose becomes of composition 
C 80 H 60 O aB , and, for the highest degree of nitration, 
C JO H 36 (NO 2 ) 1B O a . But pyrocellulose has a composi- 
tion C 39 H 38 (NO a ), a O 3B ; therefore, the number of inde- 
pendent nitro-celluloses (nitric ethers) may be very 

This is very important in the conception that the 
nitration of cellulose may be carried up to any desired 
degree, and for known concentrations a mixture of 
nitric and sulphuric acids neither dissolves nor reacts 
upon a product of nitration. Again, cellulose is the 
most widely disseminated in nature of all the hydro- 
carbons possessing the composition C.H 10 O B , consti- 
tuting the tissues of all plants and prepared from time 
immemorial in great masses as cotton, flax, paper, 
etc., while its products of nitration present an unal- 
terable material suitable for conversion into smokeless 


powder. This side of the matter needs no further 
elucidation, but it must be remembered that before 
the development of pyrocollodion, it was stated that 
the higher the nitration of the cellulose the higher the 
explosive produced, and that in manufacturing powder 
from highly explosive nitro-cellulose (of composition 
about C 30 H 36 [NOJ U O 25 ), collodion(of composition about 
C 30 H 40 [NO 3 ] 10 O 8B ) was added, for the reason that higher 
nitro-celluloses in the form of filaments or dust were 
easily detonated (hence their employment for mines), 
while the Tatter property was reduced or caused to dis- 
appear after gelatinization, of which collodion was 
easily susceptible and for which purpose it was added. 
The introduction of pyrocollodion changed existing 
views upon the subject, showing that maximum force 
for nitro-cellulose was not to be sought from the high- 
est degrees of nitration (i.e., for maximum content of 
nitrogen and oxygen), but that it obtained for that 
mean degree of nitration present in pyrocollodion. 
For the latter material F" 1000 = 81.5 ; while for nitro- 
cellulose of maximum nitration, C 8 N 7 (NO a ) s O 6 , F 1000 = 
74. 1 . The above represents only one side of the theo- 
retical investigation of materials suitable for smoke- 
less powder; but other considerations are also in ac- 
cord, as will be shown later; and we have, therefore, 
gone considerably into detail, the more urgently since 
it has been necessary to struggle with prejudice, 
harmful to success in such a new field as that of 
smokeless powders. 

Among the possible materials proposed, apart from 
mixtures of such different bodies as ammonium nitrate 


and various organic substances (such mixtures were re- 
jected in practice), must be considered the nitro-com- 
pounds corresponding to benzol and its derivatives, 
naphtalin, etc., as coal-tar constitutes an abundant 
source for their production in large quantities, and 
they are easily nitrated. From the class of the so- 
called "aromatic compounds" derived from benzol, 
C 6 H 6 , it is useless to expect smokeless explosives de- 
veloping large volumes of F" 1000 , although many are 
high explosives, beginning with Melinite or picric 
acid, C 6 H 9 (NO a ) 8 OH, which constitutes a powerful 
material, although far from the best, for torpedoes 
and explosive shells; and since some of the first 
smokeless powders were mixtures containing picric 
acid. The cause of the small gas volumes F IOOO devel- 
oped by the aromatic compounds is due to their com- 
position, as they are all low in hydrogen. This view 
may be illustrated by a few examples. 

To benzol there correspond bodies the general com- 
position of which may be expressed by the formula 
C a H,_ a (NO a V If a equals 1,2 or 3 (these substances 
are known and easily obtained), the oxygen content is 
insufficient to consume the carbon into CO and the 
hydrogen into H,O, although explosion occurs with 
the formation of carbon (smoke, soot) and of hydro- 

Total smokelessness could be realized from mixtures 
of highly nitrated products, as : 

4 C.H 3 (NO a ) 4 + C.H 4 (NO,), = 3 oCO +6H 9 O+ 9 N a . 
Mol. \vt., 4X258+168 = 30X28 + 6X18+9X28=1200. 
Volume of gases = 30X2 +6X2 +9X2 =90. 


If, instead of such non-existing highly nitrated ben- 
zols, pyroxylin be employed (as in the American 
smokeless powders of Munroe and other inventors), a 
gas volume approaching that developed by pyro- 
collodion is realized : 

6C 9 HNO a + 26CHTN 9 11 = I 9 2CO + io6H a O 
Mol. wt., 6 X 123 + 26 X 297 = 192 X 28 -f 106 X 18+42X28 = 8460. 
Volume of gases = 193 X 2 -f 106 X 2 +42X2 =680. 
Fiooo = 80.4. 

The matter assumes a different aspect if ammonium 
nitrate, NH 4 NO 9 , be employed for converting the car- 
bon and hydrogen of nitro-benzol into CO and H 3 O. 
This substance contains a large quantity of hydrogen for 
a relatively small content of oxygen (but for this re- 
sult a large quantity of NH 4 NO 3 must enter into the 
mixture), and greatly increases the volume of gas 
developed, as is evident from the following : 

2C,H.NO a + I3NH 4 N0 3 = I2CO 
Mol. wt., 2 X 123 + 13 X 80 = 12 X 28 -f 3i X 18 -f 14X28 = 1286. 
Volume of gases 12 X 2 +31X2 +14X2 =114. 

But mixtures of this salt must always be avoided if 
a satisfactory smokeless powder is to be produced, as 
it is soluble in water, as well as hygroscopic, and pro- 
duces with viscous, oily materials only coarse mechan- 
ical mixtures. 

Similar results are to be obtained from other aro- 
matic substances, and we may refer by way of ex- 
ample to mixtures of pyroxylin with picric acid, 
C.H/NOO.OH, and nitro-naphtalin, C 10 H 7 (NO 3 ), 
such compounds having been recently experimented 


with (but abandoned in practice) as smokeless pow- 
ders. Among other disadvantages, they develop 
smaller gas volumes, F ]000 , than pyrocollodion, on 
account of their relatively small content of hydrogen. 

After an examination in the above manner of the 
composition and properties of all possible materials 
capable of employment as smokeless powders, we 
arrive at the following deductions in relation to the 
volume of gases (measured at given temperature and 
pressure), F" 1000 , developed by their combustion: 

I. Only substances containing nitrogen, carbon, 
hydrogen and oxygen are capable of entire conversion, 
as required for smokeless powders, into gases that do 
not react upon gun-metals. Hence all other explo- 
sive substances (e.g., fulminate of mercury, chloride of 
nitrogen, etc.) containing haloids, metals, phosphorus, 
etc., are unsuitable for use in gunpowders. 

2. When the combustion of carbon results in the 
formation of carbonic acid gas, CO a , a less volume of 
gas is formed than when carbonic oxide, CO, is the 
resultant product ; and as the former requires more 
oxygen than the latter, the increase of oxygen or 
nitrogen (if, as is usually the case, the oxygen enters 
into combination with the aid of the elements of nitric 
acid) is injurious, instead of useful, although there 
exists full conversion into gases as is required for 
smokeless powder. 

3. The greater the quantity of hydrogen in the 
powder, other conditions being equal, the greater the 
gas volume, F 1000 , corresponding to the combustion of 
the powder; and, therefore, substances derived from 


the limiting (saturated) series of hydrocarbons are 
more suitable than bodies of the "aromatic" series 
for smokeless powders. 

4. Not any of these explosive materials not contain- 
ing carbon (as N t H, NH 4 NO a ), that evolve large vol- 
umes of gas P 1000 and decompose upon ignition, are 
such as will not detonate, i.e., evolve their gases so 
rapidly that they crush the walls of guns ; whence it is 
useless to consider them as materials adaptable for 
conversion into smokeless powders. 

5. Some of the materials containing but little car- 
bon and much hydrogen may prove suitable for use as 
powders or powder mixtures, evolving large volumes 
of gas upon combustion ; but, so far as known, they 
are either volatile, or liable to decompose spontaneously 
and detonate, or else they are prepared with difficulty 
from mixtures not widely disseminated, so that at 
present it is useless to look for materials for smokeless 
powder from among them. 

6. Nitro-glycerin itself develops but a small gas 
volume (F 1000 = 63.9), as it contains an excess of oxy- 
gen. It may be employed in mixtures to form smoke- 
less powder, and its mixtures with nitro-cellulose, such 
as Cordite and Ballistite, which are practically homo- 
geneous from a physical standpoint, develop gas 
volumes, F 1IOO , a little less than that evolved by pyro- 
collodion (although such mixtures erode guns, as 
already stated). 

7. Cellulose, C 6M H 10M O BW , is a substance widely dis- 
seminated in nature and of general industrial employ- 
ment; by its non-volatility, insolubility, durability, 


etc., and by the readiness with which it is nitrated (as 
it contains much hydrogen), it constitutes a superior 
base for smokeless powders. 

8. Among all the forms of nitro-cellulose capable of 
smokeless combustion, the maximum gas volume, F 1000 , 
corresponds to C 30 H S8 N la O 49 (= 12.44 P er cent, nitro- 
gen), which is pyrocollodion, for which F 1000 = 81.5, 
and which is capable of complete gelatinization in a 
mixture of ether and alcohol, in which form it is com- 
pletely free from any tendency to detonate. In the 
first place, it is the most suitable of all the nitro- 
celluloses; in the second, it is the most rational and 
readily obtainable form of smokeless powder, destined 
to supplant not only other smokeless powders, but also 
to replace, by reason of its greater homogeneity and 
its combination of qualities, other pyroxylin powders. 

Pyroxylin powder is a mixture of nitro-celluloses, 
of higher nitration, such as C 6 H,(NO a ) 3 O 6 , and of lower, 
as C e H 8 (NO 2 ) 2 O 6 ; pyrocollodion is a definite homo- 
geneous single form of nitrocellulose. By changing 
the proportional relation of contents of the high, or 
insoluble, and low, or soluble, nitro-cellulose, it is 
evidently possible to make the pyroxylin approach 
the pyrocollodion powders; but (as has been shown 
in recent years, especially in cannon-powders) the limit 
of improvement of these forms always falls short of 
pyrocollodion. The latter is homogeneous and un- 
changeable, while pyroxylin powders vary according 
to their composition. However, from its origin it is 
in no wise different from the perfected powder of 
Vieille (although considerably different from the 


original form thereof), presenting instead of a mix- 
ture, from the chemical and mechanical standpoint, 
a homogeneous limiting mass of the composition 
C, H g8 N 12 O 49 , which is required in order that the pow- 
der may create upon combustion the maximum volume 
of vapor and gases. It is certain that henceforth 
pyroxylin powder will continue to approximate to 
the pyrocollodion until the two become identical. In 
brief, pyrocollodion represents the Russian limit of 
modification and improvement of the French pyroxy- 
lin powders, the development of which marked an 
epoch in ordnance progress, but which has not hitherto 
presented an invariable and constant relation of the 
elements entering into its composition. In this light 
pyrocollodion powder may well be styled Franco- 
Russian. Begun in France, it has been completed in 



WITHIN recent years different varieties of nitro- 
cellulose which hitherto have possessed only very 
restricted uses have had their fields of application 
broadened, especially in their relation to the manu- 
facture of explosives. 

Photographic collodions, celluloid, artificial silk, 
etc., have nitro-cellulose for their bases. On the 
other hand, besides gun-cotton, employed for military 
purposes, as for the charging of torpedoes and other 
similar appliances, nitro-cellulose enters into the com- 
position of various kinds of smokeless powders, gum- 
dynamites, cartridges for use in presence of fire-damp, 

Each of these applications demands, so to speak, a 
special variety of nitro-cellulose, and the government 
explosive-factories entrusted with their preparation, 
so far as relates to their use as explosives, have been 
gradually forced to meet requirements differing widely 
as to character. 

For a long time the Moulin Blanc factory, which 
was established for the production of naval gun- 



cotton, had scarcely more to do than to produce gun- 
cotton of maximum nitration for military uses. For 
this the well-known mixture of three parts sulphuric 
acid by weight of 65.5 Baume" and one part by weight 
of nitric acid of 48 Baume, has always been employed. 

Since then the Moulin Blanc works and those at 
Angouleme, where the manufacture of gun-cotton was 
established in 1887, have had to produce other kinds 
of nitro-cellulose, which it became necessary to man- 
ufacture practically in a regular manner. 

The theoretical researches of M. Vieille, described 
in his note of September 13, 1883, inserted in Vol. II 
of the " Memorial des Poudres et Salpetres " (p. 212 
et seq.), classified different varieties of nitro-cellulose 
in the manner following, in accordance with the 
formulae for their chemical composition : 

Vol. of Nitrogen Dioxide 
Disengaged per gram 

Cellulose endecanitrate 214 ) 

v gun-cottons, 
decamtrate 203 ) 

" enneanitrate 190 \ 

" octonitrate 178 > collodions. 

" heptanitrate 162 ) 

" hexanitrate 146 \ 

l< pentanitrate 128 V friable cottons. 

" tetranitrate 108 ) 

Gun-cotton, as well as friable cottons, are insoluble 
in a mixture of alcohol and sulphuric ether, while 
collodions, on the contrary, are soluble in such a 

Indeed, it is hardly possible to isolate each one of 


these varieties, and one always finds mixtures of two 
or more neighboring varieties. 

Thus in practice, the products obtained are classed, 
not into cellulose endecanitrate, cellulose decanitrate, 
etc., but into gun-cotton properly so-called, for which 
the content of nitrogen is between 210 and 200 c.c., 
into superior and inferior collodions, with nitrogen 
content ranging between 190 and 1 80 c.c. 'for the 
former and from 180 to 170 for the latter; .and friable 

Gun-cotton, properly so-called, is theoretically in- 
soluble in ether-alcohol. In reality it always contains 
small quantities of less highly nitrated soluble products. 
Collodions deposit in this solvent an insoluble residue, 
due to the presence of non-attacked cotton, or friable 
cottons, or even more highly nitrated cottons. 

It is for this reason- that solubility in ether-alcohol 
is determined the same time that nitration is estimated. 

The different soluble collodions are distinguishable 
from one another, moreover, by the greater or less 
viscosity of the solution obtained. As this property 
is one that may possess practical importance in the 
employment of collodions, it appears useful to study 
it simultaneously with the solubility and the content 
of nitrogen. 

A series of researches were undertaken at the An- 
gouleme powder- works for the purpose of determining 
the practical method of obtaining a chosen product 
from the series of those just enumerated. These 
researches do not claim to possess the scientific value 
of those made in 1883 by M. Vieille upon the nitra- 


tion of cotton. The resources at the powder-works 
laboratory, especially in relation to personnel, are too 
limited to justify the hope of rigorous precision for 
results obtained. But apart from their utility as a 
guide in the manufacture of a desired product, they 
seem to possess a certain interest in relation to the 
conditions controlling the nitration of cotton. 


M. Vieille in his researches in 1883 had employed 
two methods of nitration. In one he used mixtures 
containing only nitric acid and water. In the other 
he added to these bodies variable proportions of sul- 
phuric acid. 

The latter method has been alone employed in 
practice; it is the one exclusively used in the trials 
that constitute the object of the present work. 

Apart from nitrous vapors, which are always present 
in appreciable but very feeble quantities, and foreign 
matters, which are never found in the acids employed 
except in negligable quantities, every mixture em- 
ployed for dipping cotton for the production of any 
one of the various nitro-celluloses contains the three 
following elements: monohydrated sulphuric acid, 
H 3 SO 4 ; monohydrated nitric acid, HNO 3 ; and water, 


If we desire to classify the infinite number of mix- 
tures that may be obtained from all possible combin- 
ations of the three elements, in groups giving rise 


finally to products of the same kind, it is necessary to 
have recourse to a graphic method of representation of 
these mixtures. 

In expressing the relations of the proportions of 
the two elements to the third, each mixture is de- 
fined by only two numbers. It is easy, then, by em- 
ploying one of them as an abscissa and the other as 
an ordinate, to express by the ordinary system of 
rectangular coordinates any mixture determined by a 
single point. 

The different mixtures giving rise to the same 
product will thus be grouped by zones, and a knowl- 
edge of these zones will facilitate the practical obtain- 
ing of a desired nitro-cellulose, whether new acids are 
employed to produce it or whether spent acids from a 
previous manufacture are employed for the purpose. 


We shall reconsider, in the order of ideas just in- 
dicated, the results obtained by M. Vieille in 1883, 
and mentioned in the preceding note. 

In all experiments under the second method, sul- 
phuric acid of density 1.832 was employed, corre- 
sponding, according to the special tables, to a degree 
by areometer of 65.5 Baum and to a percentage of 
water of about 8 ; and, upon the other hand, to nitric 
acid of density 1.316 which, if it be supposed free from 
all traces of. nitrous vapors, should mark 34.6 by the 
Baume areometer and contain 50$ of water. 

Applying these figures to the volumetric proper- 



tions of each mixture given in the table on page 87, 
the following new table is formed : 

No. of Ex- 

Proportions of Components 

Per cent, of Ni- 
trogen Dioxidein 
Product ob- 


H 5 S0 4 

HN0 3 

H a O 
































not deter- 
































Under the system of representation adopted, by 
which the proportions of the two other elements are 
expressed as ratios to 100 parts by weight of H,SO 4 , 
the figures of the preceding table may be graphically 
expressed as in Fig. i, page 133. 

All the points representing the mixtures employed 
lie upon the same straight line ab, since they are 
formed with the same nitric acid marking 34.6 by the 
Baume" areometer. This line is inclined at 45, since 
the acid chosen contains as much water as mono- 
hydrated acid; it starts from the point a correspond- 
ing to eight per cent., there being exactly that much 
water in one hundred parts of the sulphuric acid to 
which the one hundred parts of the other components 
are expressed as ratios. 

Outside of this line there exist an infinite number of 
other points corresponding to acid mixtures the action 


of which upon cotton was not studied by M. Vieille. 
By employing nitric acid of increasing strengths succes- 
sively, along with sulphuric acid of 65.5 Baume", mix- 
tures will be obtained represented by points upon 
the lines ac, ad, etc., which start from the common 
point a and are more or less inclined according to the 


o 45- 






5 Ct i'o 15 20 25 3b & 40 45 SO 


FIG. i 

quantity of water in the acid employed. The practical 
limit is the line af corresponding to 48 Baume", the 
strongest obtainable in current manufacture, and which 
contains about 10 per cent, water. 

These are the regions left aside in previous studies, 
the exploration of which is now undertaken from the 


standpoint of practical results to be obtained industri- 
ally from the corresponding acid mixtures. 


In this first series of experiments the base material 
was the chemically pure absorbent cotton employed in 
pharmacy. ~ The acid mixtures were prepared direct 
from the three principal elements, sulphuric acid, nitric 
acid and water. For this purpose a carboy of sul- 
phuric acid of 65.7 Baum6 was drawn from current 
supplies. On the other hand a certain quantity 
of nitric acid of the highest possible strength had 
been collected at the nitric-acid factory, and from this 
nitrous vapors were completely expelled by passing 
carbonic-acid gas through it for a number of hours at 
a temperature of about 70 C. After expulsion of 
fumes this acid marked 47.7 Baume. 

The selected sulphuric and nitric acids were kept in 
carboys carefully sealed and provided with a com- 
pressed-air emptying system, the air used being care- 
fully freed from all traces of moisture. As all the ex- 
periments could not be carried on at the same time, it 
was necessary to provide that the elements entering 
into them should be always identical. 

The water present in the acids was that shown by 
the special table corresponding to their areometric de- 
gree, this quantity being checked by direct analysis, 
and was 6.5 per cent, for the sulphuric and 10.5 per 
cent, for the nitric acid. 

The mixtures employed for the experiments were 
determined by adding to the quantities of the two 


acids taken the quantity of water necessary to bring 
the total weight up to 1.2 kilos. The percentage 
compositions were deduced from these weights and 
that of the water in the acid components. 

Each mixture was divided into three lots of 400 
grams each, of which two were employed for dupli- 
cate experiments and one was held in reserve. 

The 400 grams of mixed acid were introduced into a 
large-mouthed flask provided with an air-tight stopper, 
and an equilibrium of temperature established ; then 
the four grams of absorbent cotton were added and 
the flask violently shaken. Under these conditions 
nitration took place almost instantaneously, and it 
may be said that each fibre of cotton was instantly 
surrounded with 100 times its weight of liquid acid. 

The flask was then corked and placed under a 
stream of water maintained at a constant temperature 
of from 1 8 to 19 C., which was that of dipping, the 
cotton remaining the whole time in the acid bath. 

Two dippings were made for each mixture studied, 
one of them corresponding to reaction double in length 
of time to that of the other. The admitted durations 
of reaction were six, twelve, and twenty-four hours, 
according to the kind of nitro-cellulose that was sup- 
posed should be obtained. 

It was not considered necessary to study reactions 
of longer duration, difficult to realize in practice. In 
certain cases it was deemed sufficient to control results 
by reconducting the experiment in a bath of the same 
composition, and allowing the reaction to continue 
for forty-eight and even up to sixty hours. 


Twenty-five mixtures, corresponding pretty nearly 
to the whole region embraced by those obtainable in 
practice, were thus experimented with ; that is to say, 
by maintaining as proportions of nitric acid and water 
expressed in ratios to one hundred parts of sulphuric 
acid by weight, from 10 to 60 per cent, of the former 
and from 10 to 45 per cent, of the latter. 

In fact, for nitric acid below a certain limit, reac- 
tions became too slow. On the other hand, from a 
standpoint of economy, its proportion cannot be too 
much increased on account of its relatively high price. 
As to water, its lower limit is fixed by the strength 
of the strongest acids obtainable in current manufac- 

The twenty-five mixtures studied represent suffi- 
ciently well, then, the region that may be explored in 
practice ; they are spaced apart regularly, so that by 
the examination of the products obtained, proper ac- 
count may be taken of the phenomenon of nitration 
throughout the region explored. 

The percentage of nitrogen in each of the products 
thus obtained was determined by Schloessing's method, 
described at the close of M. Vieille's paper above re- 
ferred to. Solubility in ether-alcohol and viscosity in 
that medium were determined under the methods em- 
ployed at the Angouleme powder- works and described 
at the end of the present paper. 

The following table presents a resume of all results 
obtained from the analysis of specimens produced in 
this first series of experiments : 






, ^ 

M ^>d MCO in t- ^ 


d M TfrO ^t" M d lO 



O NO d CO d CO CO 00 


g * 

MMQoor^ MOO 

ation of F 


M T}-O comm Od\o co M ^- 











c fi 



s d- 

S^^^g^^^g: ^ o^^w^R^o^o ^oco"2 0^0 o" 

g ^ 





co r^ M co M Tfo O O co m 
d vo co ^*O d in co M d in 

M M d M M 



m 1^ CMnO 0.0 -oo IO*M n oo 






S R S^^gRSS*^ S 8 

1 " 

O d vO rf ^co ^ 't'vO vO ^ T}- O Tj- vO ^ 


S" 0^ ON^CO Ro'^^S ^Oc? 0^ 1C 


O coco d co d ^t d ONCO mM d rt-d Ttr>ii-(in>-io ONM ot^ 



O> ^CO Tf rf ^- CO Tf CO ONCO T ^tCO CO rj- ON ON *f *f O*CO ON s OO 









*3 3'D 

^ - M 




"o S 

^ Jx^^N rik^>^S ^S/*^^^ 


Inspection of this table shows that mixtures in 
which the percentage of nitric acid is low, and no- 
tably mixtures V and VI, produce after twelve hours 
reaction an incompletely nitrated product. Even 
after twenty-four hours the nitration is not com- 
plete; a fact verified by certain complementary ex- 
periments for these latter. They will be discarded, 
then, as out of the category of practical mixtures. 

The graphic representation of the other mixtures, 
by the method above indicated, is reproduced below 
(Fig. 2). By the side of each point representing one of 
them, the corresponding number is indicated in Ro- 
man numerals; and the principal qualities of the prod- 
uct taken from the preceding table are expressed in 
ordinary figures, the figures chosen for each reaction 
being those corresponding to the longest duration of 
reaction. The first number is the percentage of ni- 
trogen; the second, the solubility; the third, when 
given, the viscosity. 

If, as indicated at the beginning of this article, the 
products be divided into gun-cottons, collodions and 
friable cottons, overlooking certain small quantities 
of accompanying products that may be mingled with 
them, it will be seen that mixtures giving rise to 
products of the same kind may be grouped in quite 
distinct zones. The lines which limit these zones are 
nearly parallel to one another. They depart a little 
from the straight line and are slightly inclined to the 
co-ordinate axes; percentages of nitrogen and solu- 
bilities alone have served for tracing them. Vis- 
cosities, which refer to collodions only, are pre- 



sented only as indicators, and are far from being the 
same for all products in a given zone. 



O U 


cr Q. 
I- -J 

g 15- 




5 10 15 20 25 30 35 40 45 

FIG. 2 

Without being able to discover, in results obtained 
for these viscosities, the expression of any well-de- 
fined law, it may be remarked, however, that they are 


in general greater, according as the corresponding 
percentages of nitrogen themselves are greater. Thus 
it is to be remarked that the mean viscosity for the 
zone of lower collodions is 65% while for the higher 
collodions the mean attains I4O S . 


The preceding results were obtained, as indicated, 
with pure wadded cotton. In general this is not the 
base material employed, but spinning waste, bleached 
and freed from grease, which costs less. The fibres 
of the cotton are more or less twisted and entangled, 
and, notwithstanding the effects of carding, there 
always remain at the time of dipping small agglom- 
erations, into which the acid penetrates with diffi- 
culty. In order to estimate the influence of the base 
material employed upon results obtained, a second 
series of trials were undertaken under conditions 
similar to those of the first series, except that the 
refuse cotton from current manufacture was substi- 
tuted for the absorbent wadding. During this series 
of experiments the temperature was maintained dur- 
ing the whole time of the dipping at about from 7 to 
8 C. The acid mixtures were prepared as indicated 
above. The sulphuric acid employed marked 65.8 
Baume, and contained only 5 per cent, of water. 
The nitric acid, freed from nitrous vapors, marked 
47. i Baume and contained 1 5 per cent, of water. The 
per cent, composition of each of the twenty-five mix- 


tures employed was calculated on the basis of these 
figures, and verified for a certain number of cases by 
direct analysis. As above stated, 4 grams of cotton 
were immersed in 400 grams of acid. The absorp- 
tion of the cotton by the liquid proceeded as rapidly 
as with the wadding. 

The table on page 142 presents the results ob- 
tained in the second series of experiments. 

More clearly than in the preceding series, it is to 
be remarked that for a certain number of mixtures, 
especially for those in which the relative proportion 
of nitric acid is low, prolongation of the reaction in- 
creases the percentage of nitration. We shall dis- 
card mixtures V and VI as heretofore, since the re- 
action, to be complete, would have to be prolonged 
too far. 

Figure 3 (page 143) represents graphically the 
results of the second series of experiments. 

As in the preceding series, different mixtures, pro- 
ducing products of the same kind are grouped in 
zones, and these zones are sensibly of the same form 
as those of Fig. 2; still again, but this time perhaps 
a little less clearly, however, the viscosity of the col- 
lodions appears to increase with the percentage of 


A third series of dippings was finally undertaken, 
the conditions of which were practically identical 
with those obtaining in practice. Bleached English 




14 Hours 


N t^ vo r-> O O 

M M OO Tj- r. M 

M \n M 



O co O O*> CO vO 


SS:T2 g ff 



J * R R S * 5 



Ni^ONooino ThxnenM N TfON rtM ^N 0-N tnto^ 



.S f:** 


s- M "ssa : ssas;*i? N s < &^s"-,^? 

g = 

WOOcOrt-'itN Tt Tt-CO NONNOONvOoOO 1 ^- 'tO O 


S*o ; ,1 

T^ ID o* ^t M cn cn ^ O t^ vO co T^CO M co o ^^ o^oo O r^* ^ O*O 




\n MO oo i-i r - O 


cnO * WM UM M M inCM^C*e>eooo N 




2 i" 

rf N oo r-vo N r^ enco Mcoc^Moooo^'t r^ 

g|' S 



t->r^covO ONQO tr)N moo M 'tr^M eno N N r>.Tfir>Ti-c<covO 




O ^tco enw w N tncnO'ON't cnoo ^ m o o^ 't "too oo co t~ r-~ 






\f) \f) I^N, ao ^oo oo oo co o ^ *O J""* r^vO *^ *^> ^^O vO r^ r** oo co cn 

t w 







'o S 




spun-cotton waste from the powder-factory supplies, 
was employed as the base material. This was dipped 
in different acid mixtures formed, as indicated above, 



5 so 





< Q 
(. > 

> X 



E 151 


S 10 

xvr .xi 

5 iO 15 20 26 30 35 40 45 


FIG. 3 

from distilled water and the sulphuric and nitric acids 
of which part had been employed in the preceding 
experiments. The proportions for dippings were al- 


ways the same; 4 grams of cotton to 400 grams of 
mixed acids; but the duration of immersion of the 
cotton in the acid bath was reduced to eight minutes. 
After such immersion at a constant temperature (12 
to 13 C. approx.) the cotton was removed, drained, 
lightly pressed, and placed in an earthenware crock 
in which the reaction completed itself, the pot being 
placed in a stream of water at the current tempera- 

The results from this third series of experiments 
are given in the table on page 145. 

These results may be expressed graphically in the 
manner shown in Fig. 4, page 146, neglecting mix- 
tures V, VI and XXV, which only produced partially 
nitrated products. 

Still again, the products obtained allow the mix- 
tures to be grouped into zones analogous to those 
resulting from the two other series of experiments. 
In each of these zones viscosities are variable and ap- 
pear not to follow a well-defined law, although high 
viscosities accompany the highest percentages of 


The experiments of which a resume has just been 
given exhibit the influence of acid mixtures upon the 
products obtained. Here, indeed, lies the principal 
element entering into the whole manufacture of nitro- 
cellulose. But other, secondary causes may equally 
influence final results, and it has appeared useful to 



vr> -<f M M N 

Tt t^ XO t^ t^ 






"3S G 


to N O CO Tt- O CO OO CO CO t^ rt- W 
W r^ M invO O O M M M O xo t^\O 

MM o o^o M o> c* co o^ r^ 

action of 

g c 

!o"u : 


N O CO O O N CO CO N 00 O N O O 

xn <r r>i ex r->- r^ <o N co N ^ o NI^ 
Ocn^- t^r-^vnTt-i-r-' 10 \o co oco 


ion of Re 



O /">OOOMO u^MCJO^r^ 
O ^OQ>'^'^ rj-ioMmo 

C4 M MM 


3 >>>* 






r^wNioOOO- r^xn rt-vo -> TJ-MVOI^^M M coo M xoco 

om MMO> ocr>o o^o M o o en c^co o w vo N 


1 -2 




o2 ' 
2 Q o 

r> >r o coco N N Z O O "1- N -^-O cnco O >r> o coco N M w xn 
O O O O N -3-cQ c>co o r^ vo rf N r^co to o^vO O oo O O r- 


g * 




xn N t^o o co O co co r>i 




3 >>-^ 

O ^vO N co M r^ Tf en xnoo O O TJ-CO O^OTt^xnOMt> 




C>Nxr)<7>OO inor^eMr^OMTfvOOMvOxrjTfOxnvO 
CMn MMCO O^coa' 1 ^-M CTC7>e>4C^co OWOM 


NOcocoOOrJ- NO TtO Tt- TtO MONO -too Tfco Tt O 

u^ " 

voxnoxnexcoco vOTt-^-McnTt-MT^-vnior^cot^Oc^Oco 

z^" 1 "" 


co co co M s coco w O coco o coo *^u - >cO'3-r > u">^tnM OO 







O CT>N COM ^-OOcomo < >Oxo too M M ^- vn T t^-oo co M CT> 

ition < 

.- u 



31 s 






^^^^^^S^^^^l^^^ > ^ > ^!SI^^^l3l^^t > ^ > 

| ( ^> ^ H-l ^-H rS fc. JL ^ ^H K> ^ ^ bj l"-^ K> rS L j* . i ^"^ b^ 

> > K ,K RX xS| Rk SggS 



study them also. Among these causes, duration of 
reaction has been briefly touched upon already, as 
well as the nature of the base material. To these must 

v 5 10 1 J 5 20 ~3T 30 35 40 45 


FIG. 4 

be added temperature during dipping and reaction, as 
well as the later effect of the manipulations that ni- 
trated cotton may undergo after the operation of 


dipping. These different effects will be examined 

Base materials. Cotton wadding nitrates more 
readily than spun-cotton waste, which nitrates with 
the greater difficulty the more it is tangled, the 
coarser its threads, and the more knots it con- 
tains. On the other hand, it equally appears that the 
previous preparation of the waste destined for dip- 
ping exercises a very sensible influence upon the vis- 
cosity of the collodions obtained. Thus it is that two 
cottons of absolutely different origin, submitted un- 
der identical conditions to the action of a common 
acid mixture, produce collodions possessing the same 
mean percentage of nitrogen, but with viscosities 
varying from what they ordinarily are to double this. 

Duration of reaction. The appended table, in 
which two or three durations of reaction correspond 
to each mixture, shows that for a same final product, 
the reaction should be the more prolonged the lower 
the proportion of nitric acid in the mixture. 

Some experiments were made to follow more 
closely the progress of the reaction; a resume of them 
is presented in the following table. These experi- 
ments refer to three acid mixtures from the first 
series, the numbers of which are recalled in reference 
to results obtained; the conditions of dipping and re- 
action are those of this series; the temperature of 
the reaction was about 20 C. approximately : 











c.c. NO a 


c.c. N0 a 


c.c. NO 2 


I hour 







2 hours 































2 4 







These experiments comprise the three principal 
types of nitro-cellulose studied; superior and inferior 
collodions and gun-cotton. They show that for the 
two former a maximum of nitration is hardly ob- 
tained before the end of two hours of reaction; for 
gun-cotton, on the other hand, from eight to ten 
hours are required. If the reaction be prolonged 
beyond these limits, the solubility has a tendency to 
increase. In these experiments viscosity could not 
be measured; but the preceding tables seem to show 
that no relation exists between it and duration of re- 

Temperatures of dipping and reaction. A certain 
number of dippings were made under conditions iden- 
tical with those of the second series of experiments, 
at the three very different temperatures of i, 12 
and 25 C. These temperatures were maintained 
throughout the whole duration of the reaction. 

Results obtained are grouped in the following 



Temperature during Dipping and Reaction 






12 C. 

25" C. 

~ 3 

u w 



gen as 
in c.c. 



gen as 
in c.c. 



gen as 
in c.c. 

in * 


























































3 6 




Other experiments were made to the same end 
with wadded cotton under the conditions of the first 
series. The results were as follows: 

Temperature during Dipping and Reaction 



c - 



ti o 

, 2C. 

15 C. 

26 C. 

VM 3 




in c.c. 



in c.c. 



in c.c. 








































































These experiments seemed to show that, so far as 
collodions were concerned, increase of temperature 
during dipping and reaction increases the percentage 
of nitrogen and the solubility, but sensibly diminishes 


the viscosity. In other words, a low temperature re- 
tards the reaction. In what relates to gun-cottons, 
the influence manifests itself less distinctly; it is ad- 
mitted, however, that at high temperatures the solu- 
bility has a tendency to increase. 

Subsequent manipulations. The different cellu- 
loses are submitted before use to a number of 
manipulations; first to washings, which are necessary 
to remove the last traces of acidity, and for certain 
ones to pulping, which serves to facilitate subsequent 

Washing is effected by a more or less prolonged 
boiling, either in pure water or in an alkaline solu- 
tion; the pulping by beating-engines similar to the ap- 
paratus employed in the manufacture of paper. 

These operations, which are absolutely mechanical, 
have no influence upon the chemical composition of 
the final product, and therefore upon the percentage 
of nitrogen; but it is different for solubility and vis- 
cosity, which are purely physical properties. 

Experiments were made with the view of ascertain- 
ing how various kinds of nitro-celluloses acted during 
these operations. Results are grouped in the tables 
on page 151. 

These various experiments were made with quite 
large quantities of gun-cotton, consequently more or 
less homogeneous, so that taking of samples should 
suffice to explain certain anomalies to be noticed in 
the progress of phenomena observed. From the 
figures in the preceding tables it may be concluded, 
on the one hand, that the two operations of washing 





i s 







oO u 

h^; a 







3 c 


























5^ s 





















be y 







5 c 










oO o 

z .s 























o O u 


*- Z C 












So 2 


~Z a 











is G 


















C '"* 



















hJs c 



























hi; e* o 

2g J 


u^ 1 


^ O. M 

< 3 

e* t>. o co 




f " 


O . 






r^ N >r> TJ- 




E m 

N W 







O . 

^- O o M 

trice CO M 


T ) 

C4 CO M 







O^ N O 




Q\ O ^t" CO 


E* 1 

CO OO M d 





Tj- \f) U^QO 


o . 

d vO >-i N 


E" 5 





Tf* V> COCO 


U-> I-H CO k-< 


^00 M M 


CO ^> rfO 

O M rr co 

r-oo M M 

u t) 



M W tO t 


in warm water and pulping have the effect of increas- 
ing the solubility of gun-cottoils; on the other hand, 
that these two operations diminish to a marked de- 
gree the viscosity of the collodions. 


As already stated, the results of the experiments 
just described complete in a certain sense those ob- 
tained by M. Vieille in 1883, since they relate to a 
series of mixtures that had not been studied up to 
this time. 

Apart from the theoretical interest they may pos- 
sess, they also possess a practical utility. 

When it is desired to produce a certain definite 
nitro-cellulose, the first thing to determine upon is 
the composition of the dipping mixture to be em- 
ployed. Heretofore, the proportions of the different 
elements, whether new or spent acids, were calcu- 
lated somewhat arbitrarily. The experiments, of 
which a resume has just been given, permit a more 
methodical procedure in such a case. 

These are only laboratory experiments, it is true; 
and in practice a thousand causes, more or less well 
understood, arise to influence final results. Some 
supplementary experiments have been made to as- 
certain how these known causes tend; but it would 
be rash to assert that it would be possible, on the 
strength of the data afforded by these experiments, 
to obtain with certainty a product of which all the 
qualities were predetermined, 


However, the great similarity of the results ob- 
tained in the three series of experiments, each of 
which approaches more nearly than the preceding 
to conditions obtaining in practice, justifies the be- 
lief that, although no one of them may enable us to 
calculate the proportions of a mixture capable of pro- 
ducing a certain product, nevertheless, the general 
trend of the phenomenon of nitration is that indi- 
cated by the position and relative importance of the 
different zones above referred to. 

The knowledge of this progress in the phenomenon 
of nitration permits us, then, after a preliminary trial, 
which, besides, is not undertaken by chance, to 
modify, if necessary, the proportions of the mixture 
first taken, in such a way as to arrive at a desired end. 

Different considerations serve as guides in the 
preparation of the preliminary mixture. While still 
keeping within the zone corresponding to the prod- 
uct desired, the relative proportions of nitric acid 
and water may be varied between quite wide limits. 
It is desirable, from the standpoint of economy, to 
diminish the quantity of the most expensive element; 
that 'is to say, the nitric acid; in certain cases, how- 
ever, to develop certain qualities in the final product, 
we may be led to increase it. Finally, if spent acids 
of known composition are to be employed, one mix- 
ture may be found more advantageous than another, 
according to the relative quantities of acids to be 

The acids used, nitric acid or spent acids, always 
contain nitrous vapors to a small extent. These 


nitrous vapors may, if present in sensible quantity, 
falsify conditions upon which it was thought reliance 
could be placed, and this fact must be borne in mind 
in calculating the trial mixture. But in practice the 
proportion does not exceed from i to 2 per cent., and 
within these limits it may be admitted that the pres- 
ence of nitrous vapors does not change results sen- 

It is necessary, besides, to bear in mind that on ac- 
count of the reaction that takes place in the dipping- 
vats, and which produces water, the proportions of 
sulphuric acid, nitric acid and water in the vats are 
no longer the same as those of the original mixture. 
These alone are the conditions which affect the de- 
gree of nitration and are determined beforehand ac- 
cording to the considerations just indicated. The 
others, which must also be known, since they serve 
in preparing the mixtures properly so-called, are 
easily deduced by means of corrections, which de- 
pend upon the size of the vats, the relative weight of 
cotton employed each time, etc., etc.; and they are 
determined in practice, by analysis, before and after 
dipping, of a certain number of mixtures most com- 
monly employed. 

When the formula for a mixture permitting de- 
sired results to be obtained is thus found, the choice 
of raw material to be dipped is also to be thought of, 
as well as temperature of dipping, duration and tem- 
perature of the reaction, etc., etc. 

Here also, the experiments recalled above may 
serve as a guide to results to be obtained. 


There always exist sensible differences between the 
industrial manufacture of a product and the labora- 
tory process which serves as a basis for it. A thou- 
sand causes, varying from day to day, arise to modify 
the qualities of the final product. At first one pro- 
ceeds only by guess-work; it is only in the end, as 
the result of the processes followed with method and 
perseverance, that one becomes able to define with 
exactness the effect produced by each one of a num- 
ber of stated modifications. 

With this line of thought a number of experiments 
were made at the Angouleme laboratory upon the 
production of nitro-celluloses. Much still remains 
to be done, but the results obtained already permit a 
more methodical procedure than in the past. 


By Lieutenant JOHN B. BERNADOU, U. S. Navy 

THE systematic development of improved ballistic 
properties from progressive explosives constitutes 
one of the most important ordnance problems of the 
present day. The idea is beginning to gain ground 
among us that hereafter we must look to the powder, 
as well as to the gun, in our efforts to increase the 
rapidity of flight and the penetrative power of pro- 
jectiles; that we must consider the source of energy 
at our disposal conjointly with the apparatus whose 
function it is to convert that energy into useful work. 

So long as the art of powder-making remained at 
a standstill as it practically did for several centuries 
while the practices of alchemy, rather than the prin- 
ciples of chemistry, may be said to have controlled 
the manufacture of all explosives, the best that could 
be done was to follow the progress of mechanics in 
efforts to effect ordnance improvement; guns were 
built and powders were found to fire from them. To- 

* Abstract of lecture delivered before the U. S. Naval War 
College, July 20, 1897. 



day, however, not only is the composition of powder 
undergoing modification, but new explosive com- 
pounds, the development of which is based upon 
chemical discovery, are coming into general use; the 
results of investigations into the chemical and physical 
properties of explosives, systematized and coordi- 
nated by the methods of mathematical analysis, have 
so increased our knowledge of ballistics, that de- 
signers of ordnance are forced to accept new condi- 
tions as factors of prime importance in the attain- 
ment of ballistic effect. 

For purposes of comparison, the old forms of pow- 
der, such as black gunpowder, may be regarded as 
imperfect mechanical mixtures of particles of the ma- 
terials of which the powder is composed; the new ex- 
plosives, as very intimate mixtures of the atoms of 
those elements from the union of which into molecules 
the substance of the explosive is formed. When the 
old powder is employed as a fine dust, it burns with 
great speed and violence; when agglomerated into 
grains* it burns in a slow, progressive manner. If the 
grains of the old powders become disintegrated before 
they are completely consumed, through effects of heat 
and gas pressure developed in the bore of the gun, the 
grains crumble away; pressures become violent and 
regular progressive combustion ceases to obtain. 
Similar, but not identical, conditions exist for the 
new explosives. In the form of dust they burn with 

* By " grain " is to be understood any regular form, flat strips, 
rods, cubes, etc. 


exceeding rapidity and great violence; when all the 
particles are decomposed simultaneously, they deto- 
nate; by building them up into dense grains they may 
be made, under favorable conditions, to burn pro- 

It was obvious from the start that many advan- 
tages were to be obtained by the substitution of the 
new explosives for the old as progressive powders. 
The former burned up completely, leaving no resi- 
due. Many of them made no smoke. Other condi- 
tions being equal, these two qualities alone would 
have been sufficient to justify their general adoption. 
But other conditions were not equal. Up to a few 
years ago the fact remained that no positive, certain 
means of making nitro-explosives burn progressively 
had been found. All known precautions could be ob- 
served in the preparation; they could be built up into 
dense grains with the greatest possible care; yet, 
every now and then a charge of the powder would 
detonate; that is, instead of burning progressively, in 
accordance with the finished form of its grains, it 
would burn as the dust from which the grains were 
built up. A gun would be shattered, perhaps a life 
or two lost, and then all confidence in the new ma- 
terial would disappear, the chosen line of develop- 
ment would be abandoned; no fundamental facts 
would be left unchallenged to anchor new hopes 

Until some way could be found, then, of firing a 
nitro-compound from a gun with positive assurance 
that detonation would not occur, there could be no 


change from the old powders to the new. This as- 
surance was, however, obtained by the discovery that 
nitro-cellulose, colloided and formed into grains of 
regular size, would in all cases, if ignited in a closed 
space, burn away in a progressive manner, at a rate 
proportional to the form and dimensions of the grains 
and to the conditions of their confinement. Two 
proofs made the fact certain that colloids would not 
detonate; first, that the grains of colloid powders 
which were shot out of the gun without being com- 
pletely consumed, preserved the original shapes, in re- 
duced dimensions, of the grains of which the powder 
charge was primarily composed; second, that not tens, 
nor hundreds, but thousands of rounds of colloid pow- 
ders, fired in guns or exploded in closed vessels, de- 
veloped in every case pressures that could be shown 
to correspond rationally, in accordance with the 
theory of progressive combustion, to size and form 
of grains and to dimensions of gun-chamber or ex- 
plosion-bomb. To make these facts certain, pres- 
sures were carried up beyond the 33,ooo-lb. limit al- 
lowed for cannon; and in explosion-bombs to well 
beyond 100,000 Ibs. per square inch. 

Now, as the only certain means yet found of avoid- 
ing detonation and of assuring progressive combus- 
tion is through the colloiding of nitro-cellulose, and 
as nitro-cellulose alone can be colloided, it follows 
that we are definitely limited in our choice of ma- 
terial for progressive powders to a certain prepara- 
tion of nitro-cellulose. It is true that a number of 
substances, such as nitre-glycerin and nitrates of 


metallic bases, may be distributed in minute particles 
throughout the body of the colloid, and that im- 
munity from simultaneous detonation may be se- 
cured for the particles so distributed; but they all re- 
main uncombined in the colloid the nitrates, as the 
sand or minute shells in the body of a sponge; the 
nitro-glycerin, as the water in the pores of the 
sponge. It is desired to emphasize by this compari- 
son the fact that all the new powders, without excep- 
tion, must be built up from some form of colloid 
nitro-cellulose, whether they contain other ingredi- 
ents or not. Thus, in the case of those powders con- 
taining nitro-glycerin we may reduce the percentage 
of nitro-glycerin to zero; that is, we may eliminate 
it. If we were to remove the colloid nitro-cellulose 
from such a powder we would have remaining nitro- 
glycerin, which would detonate in the gun upon the 
attempt to fire. 

Nitro-cellulose, which is usually prepared by dip- 
ping cotton into nitric acid, possesses a property 
which the cotton, before dipping into acid, does not, 
i.e., of dissolving in a number of substances. One 
of these is acetone, a volatile fluid, with a character- 
istic pungent and aromatic odor, somewhat suggest- 
ing common alcohol in appearance and properties. 
Another.solvent for a different kind of nitro-cellulose 
is a mixture of ether and alcohol. If the clear liquids 
which constitute the solutions in these substances be 
evaporated, there will be obtained, not the nitrated 
cotton, in its original fibrous form, but first, a 
syrupy liquid, then a jelly, and finally, as dryness is 


approached, a solid translucent mass, varying in color 
according to the variety of nitro-cellulose from which 
it is prepared, from a straw-yellow to a chocolate- 
brown, and generally suggesting, in its various forms, 
tortoise-shell. To such a substance the appella- 
tion colloid, from its glue-like consistency, has been 

The general name for the material produced by 
steeping cellulose into nitric acid is nitro-cellulose. 
One of its common forms is gun-cotton. Chemists 
are well aware that there are many different kinds of 
nitro-cellulose, but just how many there are no one 
has as yet even been able to predict, the exact com- 
position of cellulose and of its nitro-derivatives re- 
maining among the yet unsolved mysteries of nature. 
Three forms were originally assigned to it, the mono-, 
di- and tri-, just as there were the three forms of 
mono-, di- and trinitro-glycerin. A later investi- 
gator (Eder) succeeded in proving the existence of 
six. The authority of to-day upon the subject, whose 
views are now generally accepted (Vieille), has form- 
ulated eight. Now, just as there are many varieties 
of nitro-cellulose, so there are many varieties of col- 
loids. The nitro-celluloses themselves all look alike; 
in their common pulped form they suggest fine white 
flour. They can be distinguished from one another 
with ease by the readiness with which some of them 
go into solution in certain solvents, while others re- 
main undissolved in these solvents like so much sand. 
The fact that they possess such different properties 
is accounted for in practice by proven differences in 


chemical composition. It suffices here to state that 
there are a number of different varieties of the sub- 
stance which form a number of different colloids. 
The question of composition will be considered later 
in relation to the gases resulting from the combus- 
tion of nitro-compounds. 

We are familiar with the kind of nitre-cellulose 
used for detonating purposes gun-cotton. We are 
also familiar with a form of colloid in common use to- 
day as a material. I refer here to celluloid, now very 
generally employed for the manufacture of a great 
number of useful articles. The nitro-cellulose from 
which celluloid is prepared may be made by steep- 
ing cotton in weak acids, and is rather a combustible 
than an explosive; it is a very different substance 
from the high explosive, gun-cotton, which is pre- 
pared from cotton by the use of strong acids. One sol- 
vent used to make celluloid is a mixture of ether and 
alcohol; the same solvent has no effect upon gun-cot- 
ton, to dissolve which acetone must be used. We 
have, then, two different types of colloid to start 
with celluloid (formed from weakly nitrated cellu- 
lose by the use of ether-alcohol), and the acetone col- 
loid of gun-cotton. It may be stated here that these 
two types of colloids represent all that is important 
in relation to colloid material for the manufacture of 
smokeless powder, as the matter has been understood 
up to a very recent date. The various colloids of the 
eight varieties of gun-cotton above referred to range 
themselves under one or the other of these two types. 
We thus have two classes of colloid to experiment 


with, as gunpowder, and all the information we pos- 
sess in relation to them is the fact that, however they 
may burn in the gun, yet they will not detonate. 

Suppose that a number of rounds of powder are 
prepared from the two colloids, how will they act 
when fired from a gun of a given calibre? Let us as- 
sume that we have at our disposal the instruments 
commonly employed for the measurement of muzzle 
velocities and of bore-pressures, the chronograph and 
pressure gauges; we will then have, as a basis of com- 
parison, first, the ratios V IP of velocities to pres- 
sures*; second, our personal observations of other 
phenomena attending explosion. 

Actual practice shows that the best results ob- 
tained for the two powders in a given gun, by vary- 
ing weights of charge and dimensions of grain, would 
be about as follows: 

Gun-cotton-acetone colloid V/P = 2100/15-19; 

Celluloid ixitro-cotton f colloided with ) v/Po-m /*& 
ether-alcohol [ y '^ 

Inspection of the results shows the existence of a 
pressure-range of from fifteen to nineteen tons for 
the acetone powder; this means that while no detona- 
tion would occur, yet that pressures would jump be- 
tween certain limits. Such a phenomenon is often 
observed in the tests of brown powders for heavy 

* A convenient expression for comparing, in a given gun, the 
ballistic properties of different progressive powders, V repre- 
senting velocities in foot-seconds, and P, pressures in tons per 
square inch. 

f Commonly called " soluble nitro-cellulose." 


guns. A powder of this character would be unsuitable 
for general use by reason of pressure irregularity. 

Upon firing the celluloid powder another and per- 
haps worse inconvenience would be met with. Con- 
siderable smoke would be developed and the interior 
of the bore would be found lined, after each round, 
with a heavy coating of soot, which, after one or two 
shots had been fired and the gun had become heated, 
would ignite after each succeeding round upon in- 
gress of fresh air on opening of the breech, thereby 
producing flaming at breech and muzzle. 

Obviously, as gunpowder, neither the one nor the 
other form of colloid is suitable. If they are to be 
employed they must be improved. Irregularity in 
pressures from the gun-cotton-acetone colloid is due 
to brittleness; if we are to use this material we must 
devise some means of toughening it. The celluloid 
does not contain enough oxygen to consume its sub- 
stance into gases; to use the latter we must put more 
oxygen into it. 

The original phases of the colloid powder question 
thus present themselves. Neither the one nor the 
other form of colloid proving suitable for direct 
manufacture into powder, people began to try to im- 
prove them by combining them, or by adding foreign 
substances to them. It was well understood that no 
form of gun-cotton, however highly nitrated, con- 
tained enough oxygen to effect its own complete 
combustion the conversion of its carbon into the 
higher oxide of carbon, CO 2 . The first idea of the 
experimenter was to add enough oxygen to the nitro- 


cellulose to thus complete its combustion, and this 
unfortunate attempt led to many years' delay in the 
development of smokeless powder. It started in- 
vestigators off upon a wrong track; complete com- 
bustion was one thing, and the work necessary to de- 
velop highest velocity with lowest bore-pressure, an- 

With the purpose, then, of improving ballistic 
qualities of powders by causing them to consume 
completely and to develop regular pressures, experi- 
menters in different countries began to try the effect 
of introducing into the colloid various foreign sub- 
stances, generally, oxidizing agents, such as ni- 
trates of metallic bases; sometimes, when the mix- 
tures became too violent in their action, a substance 
rich in carbon, called a deterrent, was added. Such 
work was a good deal like groping in the dark. 
There was no method in it. But there was one great 
incentive to keeping it up, viz., the fact thereby es- 
tablished that the addition of these nitrates to the 
colloids actually increased the velocity developed for 
a given bore-pressure, whatever the inconveniences 
attendant upon the employment of these mixtures as 
powders may have been.* 

To establish a comparison between the ballistic 
efficiencies of the two types of pure colloids above 

* The increase in muzzle velocities for a given bore-pressure to 
be attained by incorporating certain quantities of metallic ni- 
trates, nitro-glycerin, etc., into the body of the colloid, con- 
stitutes a special phase of development of progressive powders, 
which will be discussed in a subsequent paper. 


cited, and of certain" compound colloid powders into 
the substance of which metallic nitrates or other oxy- 
gen-carriers are incorporated, the tabular record of 
performances of powders of these classes is submitted 
on page 167. Note is made therein of objectionable 
features developed for each of the explosives named. 

Referring to the table we find powders A and B 
with properties as already described. - The K, BN, 
and cordite all make good ballistic showings, giving 
velocities greater by about 300 ft. sec. for a de- 
veloped pressure than the former. 

The last line of the table shows that each of the 
powders possesses certain unfavorable qualities which 
militate against its adoption for service use. The 
gun-cotton acetone colloid develops irregular pres- 
sures; the ether-alcohol colloid of soluble nitro- 
cellulose deposits soot; the K and BN produce some 
smoke and bore-deposit. Cordite contains a volatile 
liquid, nitre-glycerin, which develops great heat 
upon combustion. Now the development of highest 
velocity at lowest pressure is most important, even 
if obtained at the expense of the production of cer- 
tain partially unfavorable conditions; but there was 
a further incentive to progress at this stage of de- 
velopment the fact there had been found a form of 
pure colloid, unadulterated by admixture with other 
substances, which, while developing high velocities 
at moderate pressures, possessed the full round of 
good qualities necessary in a service powder. 

The end sought for in the development of gun- 
powder is the attainment of the capability of deliver- 


-1 s 



0. jn 





O DQ jj| - 'g o 

... g 

o *; 1) 13 TJ 



^ j V ^ ^ rt 


2'H, w 

O "o ' * 

o o 

!_ ^ M 



i K"2"e 
S ?|| 



g v, ._. ^ T? 

1 " 








c v 


o- 2 

fc, n 

.Q j, n 


D *^ 4J 4- 4J 


Q-,^*^ y *^3 *b^ tli 


>, " o "^5 i: 

8 u. 

"S|| g' 5 

. M 



J^ 4) ^ 2 


s S* 

2 S S j-j ^ " 



El- M 

c o 






"3 2 

N ^ 





*<3 ^ 



iL *^ o 



4) ,Q OT 






o . . 




w u-> 












1 ^ 



1 68 


ing most accurately in a given interval of time the 
greatest number of most powerful blows; this result 
to be effected with minimum risk to gunners and with 
least possible injury to gun. Hereby is implied the 
fulfilment of a number of important independent 
conditions, no one of which may be overlooked in the 
effort to successfully accomplish the object sought. 
These conditions correspond to qualities possessed 
by a powder, and may be given, with their opposites, 
a general grouping under headings as follows: 








Positive -j 


of minimum 

Formation of 

Good keep- 
ing quali- 


Negative \ 



Smoke and 


Low value 
of V/P 

The first condition is of paramount importance 
and limits us to the employment of a colloid material. 
The second represents the fact that the greater the 
heat developed the greater the wearing away of the 
inner surface of the bore. The third requirement 
means obviation of bore-deposit, which operates to 
reduce rapidity of fire by necessitating more or less 
frequent sponging, and to diminish accuracy of prac- 
tice through the formation of smoke. The fourth re- 
quirement, stability, is all-important, when we re- 
member that the ship must carry safely her store of 
powder for at least a cruise, and for the reason that 


when powders begin to decompose they lose their 
homogeneity and crumble which means an end to 
regular-pressure development. 

Let us now take up requirement fifth the attain- 
ment of maximum propulsive effect and see what it 
leads to. What causes the expulsion of the projec- 
tile from the gun? The expansion of powder-gases. 
What limits our employment of the expansive force 
of these gases? The attainment of the limiting bore- 
pressure, this limit being commonly set at fifteen tons 
per square inch. What represents the greatest 
amount of work in the form of velocity? The great- 
est amount of gas-expansion in the gun behind the 

We desire, then, the greatest amount of gas ex- 
pansion, but we are limited as to the rate of this ex- 
pansion; that is, we must not let it develop in the 
bore a pressure greater than fifteen tons in the gun- 
chamber or upon the interior walls of the piece. This 
is tantamount to saying that to produce maximum 
velocity we require the evolution, at a suitable lim- 
ited rate of expansion, of the greatest possible 
volume of gas, the expansion of the gas constituting 
under these conditions the propelling impulse. 

It may be stated here that we possess, within limits, 
the power of controlling the rate of combustion of 
colloid powders -by varying certain conditions re- 
lating to combustion, the principal of which are (i) 
the size of the grains of which the charge is com- 
posed, (2) the volume of the powder-chamber, (3) 
the length of the bore of the gun, (4) the weight of 


the projectile. Granting, then, that we possess 
within limits the capability of controlling the rate of 
evolution of the powder-gases, we are led to the fol- 
lowing conclusion: that the best smokeless powder 
is that stable colloid which, for a given weight of its 
substance, evolves in the bore of the gun at the most 
suitable rate of evolution and expansion, the greatest 
volume of gas, the said evolution being accompanied 
with the development of the least heat. 

We are led by this deduction to regard the action 
of the powder from a new standpoint. Besides con- 
sidering what a powder is composed of, w r e must now 
consider what gases it is converted into upon decom- 
position, and what volumes of these gases it gener- 
ates. Most important of all, it leads us to conduct 
experiments for the purpose of ascertaining what 
colloid will, for a given weight of its substance, liber- 
ate the greatest volume of gases. 

Here was one starting point for a series of experi- 
mental researches culminating in the discovery of an 
efficient smokeless powder. Another line of experi- 
mental approaches to the same end connected the 
efforts to toughen the substance of colloid films con- 
taining sufficient oxygen to effect their own com- 
plete combustion, with a view to the attainment of 
increased regularity in developed pressures; a third 
related to the determination of the causes of certain 
ballistic phenomena hitherto unexplained, e.g., that 
acetone colloids developed, in some cases, greatly 
improved ballistic qualities, after the lapses of periods 
of from six to twelve months from time of manufac- 


ture. These several lines of experimental investiga- 
tion proved in the end to converge towards the at- 
tainment of a common result the development of a 
special form of nitro-colloid that possessed the 
toughness and therefore the regularity of burning of 
celluloid, and that contained enough oxygen to con- 
vert its substance, upon ignition, into wholly gaseous 
products which the celluloid did not and that 
liberated not only the greatest volume of powder- 
gases at the most suitable rate, but the greatest 
volume of gases that can be evolved by any colloid 
at any rate of combustion, whether these colloids 
contained or did not contain nitro-glycerin, metallic 
nitrates or other substances. 

The questions now present themselves: in what 
ways do we possess the control of kind and amount 
of gases evolved upon the combustion of nitro- 
cellulose and of its colloids; how does the constitu- 
tion of these gases vary; what are they? 

It has been already stated 'that the exact chemical 
formulae for cellulose and its nitro-derivatives have 
not yet been written. That for cellulose may be ap- 
proximately expressed as C 6n U lon O 5H , where n is an 
undetermined or indeterminate numerical quantity. 
When cellulose is steeped in nitric acid or in a mix- 
ture of nitric and sulphuric acids, it is converted into 
the substance the composition of which may be ex- 
pressed asasC 6w H IOW _^O s (NO 2 )^. These two expres- 
sions, C 6w H IOW O 5 and C 6w H IO . rtW O 5W (NO 2 ) aw , may be 
considered in comparison with one another. But 
contain the same quantity of carbon; the quanti- 


ties present of the other elements are changed by 
nitration; an atoms of hydrogen are displaced by 
an equivalents of a combination of nitrogen and 
oxygen (NO 2 ). This additional oxygen from the 
NO 2 acts to supply the energy that converts the 
cellulose into an explosive, and enters into its sub- 
stance in combination with a certain quantity of nitro- 
gen. Why it carries the nitrogen with it we do not 
know, but it is a chemical fact that it does so the 
fact upon which the designation of the new ex- 
plosives as nitro-derivatives or nitro-explosives is 

Upon the ignition of the nitro-cellulose or its 
colloid the nitrogen is set free; the hydrogen com- 
bines with its equivalent of oxygen and appears in 
the air as steam; the remaining oxygen unites with 
the carbon to form gaseous oxides of carbon. If 
there be not enough oxygen to consume all the car- 
bon into gas, part of the latter is deposited as soot; 
this result was obtained in the attempt to employ 
celluloid as gunpowder. If there is enough oxygen 
to consume all the carbon into gases, we have, as 
products of combustion, a mixture of the gaseous 
oxides of carbon, of which there are two, CO 2 and 

The property possessed by carbon of combining 
at high temperatures with intense energy with 
oxygen, to form gaseous oxides, is the fact upon 
which the practical development of explosives de- 
pends; there are many explosives that contain neither 


carbon nor oxygen, but with these we yet no 
practical relations in ordnance matters. 

It has been stated already that the attempt to ob- 
tain what is called complete combustion for nitro- 
cellulose colloids by incorporating oxidizing agents 
into them had misled investigators, who confounded 
the attainment of complete combustion with the de- 
velopment of maximum velocity at lowest pressure. 
" Complete combustion " means the conversion of all 
the carbon into the higher oxide of carbon car- 
bonic acid, CO 2 a dense gas about 1.9 times as 
heavy as the air, the formation of which is accom- 
panied with the development of a high degree of heat. 
The complete combustion of carbon into carbonic 
acid gas with the corresponding evolution of a great 
amount of heat is the characteristic of the combus- 
tion of nitro-glycerin. The lower oxide of carbon, 
CO, is a gas much less dense than carbonic acid, pos- 
sessing a density of about 1.4 times that of air. Sup- 
pose that we have a given weight of a compound of 
carbon and oxygen, with the elements taken in such 
proportions as to produce, on ignition, complete 
combustion into carbonic acid gas, CO 2 , with the ac- 
companying evolution of a large amount of heat; 
let us also suppose that we have an equal weight of 
compound of the same elements in such proportions 
as to develop, on ignition, the lower oxide of carbon, 
carbonic oxide, CO; then, the latter compound, de- 
veloping the lower oxide, would liberate a volume of 
gas nearly 1.9/1.4=1.36 times greater than the 
former. The greater heat produced by the forma- 


tion of the carbonic acid gas would cause the volume 
of that gas evolved to be the more expanded, but, 
at the same time, and this is a fact of crucial impor- 
tance in the present work, the greater heat would 
cause the gas to be generated at a more rapid rate, 
in fact, at an extremely rapid rate; and what we re- 
quire is a low, regular rate of gas evolution to pre- 
vent our exceeding at any time the set limit of per- 
missible bore-pressure. 

In our effort to generate from colloid nitro-cellu- 
lose the greatest volume of gas at the most gradual 
rate of expansion, we must seek, then, (i) to avoid 
the formation of CO 2 when CO may be formed in 
lieu thereof; (2) to avoid the formation of free car- 
bon; and (3) to generate the maximum volume of the 
lower carbon oxide. If we give due consideration 
to the amounts of water and of nitrogen formed 
simultaneously with the oxides of carbon, we shall 
find that the form of nitro-cellulose developing the 
greatest volume of gas at the most suitable rate, cor- 
responds to the formula C 30 H 38 (NO 2 ) 12 O 2 5, which 
breaks up on decomposition into 30 CO + 19 
H 2 O + 12 N. 

This material is a new type of nitro-cellulose,* de- 
veloped by experiment to meet ballistic require- 

* This special form of nitro-cellulose, which corresponds to a 
content of nitrogen of 12.44 per cent, and which was first devel- 
oped in Russia by the eminent chemist, Professor D. Mendeleef, 
has been independently developed at the Torpedo Station, 
through the study of effects of variation of times of immersion, 
temperatures of nitration and of washing, and strength of acids 
employed in the nitration of cellulose. 


merits, which contains just enough oxygen to convert 
its substance into a gaseous body. Its formation 
from cellulose depends upon strengths of nitric and 
sulphuric acid, temperatures of reaction and time of 
immersion of the cellulose in the acids from which the 
material is prepared. With ether-alcohol it forms a 
colloid that possesses, on the one hand, the tough- 
ness, and therefore the capability of development of 
regular pressures, of celluloid; and on the other the 
capability of consuming into wholly gaseous prod- 
ucts that characterizes the gun-cotton acetone 
colloid; while as a powder it develops, with present 
types of guns, excellent values of V IP of about 
2400/16. Briefly, it may be described as a celluloid 
containing enough oxygen to convert its substance 
(when it is consumed out of contact with the atmos- 
phere) wholly into gaseous products. 

It was stated in a preceding paragraph that the bal- 
listic effect produced by a progressive explosive de- 
pended directly upon the volume of gas it evolved 
upon combustion, but was not directly dependent 
upon the attainment of complete combustion. As- 
suming total conversion from solids into gases, and 
non-liability to detonation, pyrocellulose was shown 
to be the form of nitro-cellulose best adapted for con- 
version into smokeless powder. As this material con- 
tains only enough oxygen to convert its carbon into 
carbonic oxide, CO less than gun-cotton, which 
converts its carbon partly into carbonic acid gas, 
COo, and partly into carbonic oxide, CO the at- 
tainment of maximum efficiency from nitro-cellulose 


was thus shown to be accomplished through a reduc- 
tion from a maximum to a mean in the quantity of 
oxygen capable of being incorporated into nitro- 

On the other hand, it was stated that the incor- 
poration of certain quantities of oxygen-carriers 
(nitro-substitution compounds other than nitro-cel- 
luloses, such as nitro-glycerin and nitrates of metallic 
bases) into colloid nitro-cellulose, led to the attain- 
ment of an increase in initial velocity of projectile for 
a given developed bore-pressure. As nitro-glycerin 
furnishes a surplus of free oxygen to aid in complet- 
ing the combustion of the gases from the nitro-cellu- 
lose, while the nitrates surrender oxygen on applica- 
tion of heat, it would appear in this case that the at- 
tainment of a more complete combustion led to im- 
provement in ballistic effect. 

We are thus brought face to face with a seeming 
contradiction how, on the one hand, we must re- 
move oxygen; how, on the other, we must add oxy- 
gen to a progressive explosive, in order to obtain 
maximum ballistic effect therefrom. In order to 
reconcile these apparently contradictory statements 
we must consider the manner of decomposition of the 
explosive in both cases. 

One chief characteristic of pyrocellulose is its 
homogeneity. It represents no mixture of explosives 
and combustibles, such as are presented by other 
forms of powders, and it is converted directly by com- 
bustion into a set of gaseous decomposition products 
that may not be varied in amount and kind. Under 


these conditions the ballistic effect of the expanding 
gases from pyrocellulose may be referred to quantity 
of charge, area of ignition-surface, weight of projec- 
tile, calibre of gun, and volume of powder-chamber. 
Other conditions affecting developed pressure and 
velocity are bore-friction and resistance of the pro- 
jectile to rotation through inertia. 

The gun may be regarded as a gas-engine in which 
the walls of the chamber and bore form the cylinder; 
the projectile, the piston. The expanding powder- 
gases perform work by imparting velocity to the pro- 
jectile, the inertia of which they overcome just as gas 
by its expansion in the cylinder overcomes the inertia 
of the piston and the parts linked thereto. In the en- 
gine the gas is admitted alternately, first at one end of 
the cylinder and then at the other; in the gun it is ad- 
mitted in rear of the projectile but once, so that the 
gun is an engine of a single stroke. In the engine 
the steam is admitted into the cylinder through a 
valve, and, after the lapse of a period of time less than 
that required for a full stroke, admission is cut off 
and wor.k for the rest of the stroke is performed ex- 
pansively; in the gun the charge of powder consti- 
tutes both the gas itself and the valve that admits the 
gas for each grain of powder may be considered as 
a notch of opening of a valve; the more grains there 
are the greater the ignition surface, the greater the 
rate of emission of gas, or the greater the number of 
notches the valve is open. 

The action of nitro-cellulose powder-gases in im- 
parting motion to the projectile is that of the gas in 


the engine cylinder. The decomposition products 
are evolved at a high pressure, and act to propel the 
projectile, just as the gas or vapor drives the piston 
in an engine. Thus far the two cases are in parallel; 
they differ in that the space occupied by the gas in 
the gun is constantly increasing, both through the 
effect of the motion of the projectile along the bore 
and from the increase of chamber space due to the 
melting away of the powder charge, while space in 
the engine cylinder is increased through the motion 
of the piston and through connection with the valve 
before cut-off. As shown in Table I, the ballistic 
value of gun-cotton colloided in acetone (for a given 

2 100 
gun) was ; that of colloided soluble nitro-cellu- 

lose, containing not enough oxygen to convert its 
carbon wholly into carbonic oxide, CO, was ^ . 

Under similar conditions of firing, pyrocellulose de- 

veloped a value of V/P g-. It may be urged 

that the ballistic superiority of the latter colloid as 
compared with that of the two former is not wholly 
attributable to character and volume of evolved gases, 
as the acetone colloid is brittle, and that prepared 
from soluble nitro-cellulose is somewhat brittle, and 
deficient in oxygen, while pyrocelluloid is of a tough, 
leathery consistency, capable of withstanding high 
pressures without premature disintegration. Never- 
theless, as these colloids prove inferior to the pyro- 
colloids for lower pressures of about 10 tons per 


square inch, at which the effects of brittleness are not 
perceptible, and for which they all afford pressures 
regularly proportional to develop velocities, it re- 
mains that magnitude of volume of evolved gases is 
a factor of prime importance in the attainment of bal- 
listic efficiency. 


The results of incorporating an oxidizing agent or 
oxygen-carrier into colloids merit special study. Sup- 
pose that a given nitro-cellulose be colloided and 
formed into strips of a number of definite thicknesses. 
If these strips be collected separately and dried, we 
may prepare from them series of rounds, each series 
composed of different weights of strips of some one 
thickness. If the length and breadth of the strips be 
great in relation to their thickness, we need consider 
only the latter element of dimension in relation to 
their mode of combustion.* 

* We have (Glennon, Interior Ballistics, chap. VI, pp. 59, 60) 

where y is the fractional part of the least dimension of the grain 
burned up to any moment; <p(y) the fractional part of the whole 
grain burned up to the same moment; and or, y, and ju, constants 
depending upon the form of the grain. 

If the grain be a rectangular parallelepiped with a square base, 
and the altitude as the least dimension, we have 

. 2x + X* x 9 

a = i -\- 2x, A = = 

i -j- 2x ' I 4* ft*' 

where x is the ratio of the altitude to the side of the base." 

Applying the above to the present case we find that if the alti- 


Upon firing series of rounds of the several powders 
from a given gun we obtain the following results as 
to their manner of explosive action: 

i. Strips of over a certain mean thickness will be 
only partly consumed in the bore; the unconsumed 
remnants will be projected burning from the gun, to 
be quenched in the cool outer air, where they fall un- 
consumed to the ground and may be picked up at 
various distances from the piece in front of the muz- 
zle, possessing the original form (in reduced dimen- 
sions) of the grains of which the charge was origin- 
ally composed. Such powders develop low bore pres- 
sures and afford low muzzle velocities. In point of 
work performed they are equivalent to smaller 
charges of quicker powders. It may be remarked 
that no work is done in raising the temperature of 
the unconsumed portions of the grains, for if the tem- 
perature of the latter be raised but a few degrees, the 
ignition point of the explosive is reached and its sub- 
stance would wholly disappear. 

2. Strips of under a certain mean thickness are 
totally consumed in the gun. They develop high 
pressures for low velocities. The thinner the strips 
the less the weight of charge required to develop the 

tude be considerably diminished (x approaches zero) we have the 
case of the thin plate; and that the constants approach the values 

a= i, A. = o, )JL = o, 

or 000 = y. 

But y depends alone on the thickness of the plate, therefore 
the speed of combustion of a plate is a linear function of its least 


limiting permissible pressure, on account of the 
greater initial surface presented by the thinner strips, 
which occasions a high initial gas development. 

3. A certain mean thickness of strip will be found, 
for which, at a set limit of pressure, a minimum 
weight of powder will develop the greatest velocity 
that can be developed at that pressure. If strips of 
other thicknesses develop practically identical veloci- 
ties and pressures for the same pressure limits, it will 
be by burning greater weights of powder. Such a 
powder may be designated a maximum powder, for 
the material from which it is prepared and for the gun 
from which it is fired. 

Suppose, then, that colloided gun-cotton of nitra- 
tion N= 13.3 develops in a given gun a maximum 

value V/P = , what will be the effect of incor- 
porating into such powder a certain quantity of nitro- 
glycerin, or of metallic nitrates such as barium and 
potassium nitrates? Assume that during the process 
of colloiding the requisite quantity of nitrates be uni- 
formly incorporated throughout the substance of the 
pasty mass, which is subsequently formed into strips, 
as before. For this material we shall find that the 

maximum powder develops a value of V/P 1 

as against for the pure colloid, a gain in velocity 

of 300 ft. sec. for a given pressure; in energy, f ), 

of about 30 per cent, 


If, in lieu of nitrates, we incorporate nitro-glycerin 
into the colloid, we will obtain a pasty mass that can 
be worked conveniently into the form of rods or 
cords, whence the name " cordite," applied to one of 
its best known types. Cordite, as used in England, 
consists of 

Nitro-glycerin 58 parts 

Gun-cotton 37 parts 

Vaseline 5 parts 

Such a powder, fired under the above conditions, 
develops a value of V/P = approx. 

There is one characteristic of powders, such as the 
K and the French BN, containing nitrates, to which 
attention is to be directed. The nitrates contained in 
these powders exist in them in a state of suspension; 
in an undissolved state. For the BN the microscope 
reveals minute crystalline particles uniformly dissem- 
inated throughout its mass; the barium nitrate em- 
ployed in the K powder is insoluble in the colloiding 
agent, acetone, and is also insoluble in the colloid, in 
which it is held in a state of suspension and of uni- 
form distribution. 

In the case of the nitro-glycerin powders it is 
known that the nitro-cellulose is not in true solution 
in the nitro-glycerin. In this connection the follow- 
ing quotation from an authority upon nitro-glycerin 
powders, Mr. Hudson Maxim, may be cited: 

" In the very early smokeless powders, especially 
those made of compounds of soluble pyroxylin (gun- 


cotton) and nitro-glycerin, it was supposed that the 
mtro-glycerin actually held and retained the pyroxy- 
lin in solution, but it has since been learned that the 
nitro-glycerin is held by smokeless powders, whether 
made from high- or from low-grade gun-cottons, in 
much the same manner as water is held by a sponge; 
in fact, the pyroxylin exists in smokeless powders in 
the shape of a very minute spongy substance, and the 
nitro-glycerin is held in a free state within the pores 
of this sponge." 

" It is possible even with powders containing as 
little as 25 per cent, of nitro-glycerin, to squeeze out 
the nitro-glycerin in a pure state by subjecting a piece 
of this powder to great pressure between smooth 
steel plates." 

The quantity of nitro-carrier (nitrate or nitro-sub- 
stitution compound other than nitro-cellulose) consid- 
ered necessary to the production of good ballistic re- 
sults, as exemplified in certain known powders, may 
be tabulated as follows: 


Variety of Nitrn-rarrW n^H Per ceot - of Nitro-carrier 

Powder in Given Wt. of Powder 

Cordite Nitro-glycerin 58 

Maxim Nitro-glycerin 10 to 25 

BN Barium- and Potassium nitrates 211025 

K Barium nitrate 14.25 

The composition and ballistic properties of the 
three classes of explosives pure colloids, colloids 
containing metallic nitrates, and colloids containing 
nitro-glycerin may be compared as follows; 

1 84 


Pure Colloid 





Gun-cot- "1 

Insol. nitro- 


ton, 85.00 

ton and 

cellulose, 38.67 

cerin, 58 

Soluble ni- 

soluble ! QA . 


nitro-cel-f 84 ' 25 

Soluble nitro- 


lose, 10.00 


cellulose, 33.23 

ton, 37 

Sod. carb, i.oo 

balanced J 


Vaseline, 5 

Solvent, res- 


nitrate, 18.74 

ins, etc., 4.00 

nitrate, 14.25 


T^/-\f o cciii m 


Calc. carb., 1.50 

lUldr^olU III 

nitrate, 4.54 


Calc. carb., 3. 65 

Volatile, 1.29 





Metallic Nitrate 

Metallic Nitrate 



Solid undis- 

Solid undis- 


of incor- 

solved parti- 

solved parti- 



cles, uniform- 

cles, uniform- 

held in sus- 

of oxy- 

ly distributed 

ly distributed 

pension like 




water in 


colloid ma- 

colloid ma- 














Remembering what has been said in relation to the 
ballistic performance of the varieties of powders cited, 
we are led to the following conclusion: 

That minute particles of an oxygen-carrier uniformly 
incorporated into a nitro-colloid and held in suspension 


in an undissolved state throughout the body of the latter, 
render more progressive the combustion of the nitro- 
colloid into which they are incorporated. 

For convenience of reference I shall refer here- 
after to the oxygen-carrier held in suspension in the 
colloid as the accelerator. Viewed in the light of the 
principle here enunciated, the several powders we 
have been considering are all similar variants of the 
pure colloid. The remark of the compounder, " that 
a little nitro-glycerin certainly does help the powder 
along," is now the more readily comprehensible. 

The methods commonly employed for co-ordina- 
ting natural species may be applied, by way of illus- 
tration, to the classification of the various types of 
progressive explosives, to establish their relations to 
one another, and to indicate the lines along which 
advances have been effected. 

We shall next consider how the accelerator acts to 
develop increased velocity without developing in- 
creased pressure. 

i. It has already been shown how it is possible 
with powders containing as little as 25 per cent, of 
nitro-glycerin to squeeze out the nitro-glycerin in a 
pure state by subjecting the powder to great pressure 
between smooth steel plates. 

If it be possible to extract nitro-glycerin by appli- 
cation of pressure from powder in which it is incor- 
porated, then there will be a tendency to flow in the 
direction of least pressure from the instant of ignition 
of a charge to that of its complete combustion. This 
would mean, first, flow from within outwards in the 

1 86 


sfij S'sss 

rt *52-X o.-gg-? 

2 "" 

> W 

3-3 -z .5-3 

il fe . 

&^ g 




c -- c ' 







** > j, r 

c-o'C rt 




SI'S 8 a 8,8 3 



a c u 

8-2 S 


gun-chamber, where a relatively large proportion of 
the nitre-glycerin would be consumed; second, flow 
in the direction of the windage, where the quantity of 
nitre-glycerin consumed would also be relatively 
great. Such action accounts for the rapid erosion of 
the surfaces of gun-chamber and rifled bore when 
powders containing nitro-glycerin are employed. 

2. The eminent Russian chemist, Professor D. 
Mendeleef, developer of smokeless powder in Russia, 
in a paper upon pyrocellulose powder, says: 

:< The chemical homogeneity of pyrocollodion 
plays an important part in its combustion, for there 
are many reasons for believing that upon the com- 
bustion of those physically but not chemically 
homogeneous materials, such as nitro-glycerin pow- 
der (ballistite, cordite, etc.), the nitro-glycerin is de- 
composed first, and the nitro-cellulose subsequently 
in a different layer of the powder. The experiments 
of Messrs. T. M. and P. M. Tcheltsov at the Scientific 
and Technical Laboratory show that for a given 
density of loading the composition of the gases 
evolved by nitro-glycerin powders varies according 
to the surface area of the grains (thickness of strip), 
a phenomenon not to be observed in the combustion 
of the pyrocellulose powders. There is only one ex- 
planation for this, viz., that the nitro-glycerin, which, 
possesses the higher rate of combustion (Berthelot), 
is decomposed sooner than the nitro-cellulose dis- 
solved in it. This is the reason why nitro-glycerin 
powders destroy the inner surfaces of gun-chambers 
with such rapidity." 


We conclude from the above that the nitro-glycerin 
incorporated into a colloid burns more rapidly than 
the nitro-cellulose forming the colloid. More nitro- 
glycerin is consumed with one part of the charge than 
with another. During the first period the products 
of combustion evolved in chamber and bore are 
largely those of nitro-glycerin; during the second, 
those of nitro-cellulose. 

Moreover, as both materials exist in an uncom- 
bined state, although in one of intimate admixture; 
as both decompose wholly into gases, while each con- 
tains sufficient energy to continue its own decompo- 
sition, once that decomposition is begun, there is no 
reason why the rates of the two decompositions 
should be equal; it would rather appear that each 
substance should decompose at the rate peculiar to 
itself, so far as it was able, under existing conditions 
of heat and pressure, to effect a separation of its sub- 
stance from the mixed mass of the powder. 

Conditions point, therefore, to there being two in- 
tervals in the decomposition of the charge, during one 
of which a maximum quantity of nitro-glycerin, and, 
during another, a maximum quantity of nitro-cellu- 
lose is burning. 

In what follows it is not intended to attempt more 
than to indicate the mode of progressive combustion 
as implying the superimposition of maxima and 
minima of effort. This may be represented graphic- 
ally in the present case as follows: 



The result of the combination of the conditions 
here indicated would be the imparting of a double im- 
pulse to the projectiles due to the successive occur- 
rence of two maxima of acceleration. Considered as 
to their limit of possible range, the successive im- 
pulses may occur incrementally, so that the accelera- 
tor may be expressed in the form 

where p' represents the pressure due at any instant to 
the combustion of the nitro-glycerin; />", that due to 
the nitro-cellulose. 

The projectile may be regarded as receiving a third 
impulse, resulting from the chemical combination of 
the gases evolved by the nitro-glycerin and the nitro- 
cellulose. According to the researches of Messrs. 
Macnab and Ristori (Proc. Royal Soc., vol. LVI, 
p. 8), the decomposition products of nitro-glycerin 

CO 2 CO CH 4 O 

57.6 2.7 



H 2 O. 



And from the same source we obtain the decomposi- 
tion products of nitro-cellulose (N = 13.3) as 

CO 2 CO CA 4 O H N H 2 O 

29.27 38.52 0.24 0.86 13.6 16.3 

What may be called the third impulse would repre- 
sent the combination at a high temperature of mul- 
tiples of decomposition products developed in the 

^FC0 2 CO CH 4 O H N H 2 0n 
l_57.6 2.7 18.8 20.8J 

.#[29.27 38.52 0.24 0.86 13.6 16.3] 

These phases may be indicated graphically as fol- 


Accelerated colloids of K and BN types containing 
metallic nitrates are next to be considered. We may 
assume that the nitro-colloid into which minute par- 
ticles of a nitro-carrier of this type are cemented itself 
burns in approximation to the law of decomposition 
of the colloid. This state of affairs is similar to, 


though not identical with, the preceding; in the 
former, both nitro-glycerin and nitro-cellulose are 
able to effect their own decomposition, evolving 
gases that recombine; in the latter, the nitro-cellulose 
alone possesses this property, the metallic nitrates 
surrendering their oxygen through the effect of heat 
developed during decomposition of the colloid. The 
successive reactions may be represented as follows: 



Instead of three maxima of effort there are two 
maxima and one minimum, the maxima representing 
the combustion of the nitro-cellulose and the subse- 
quent combination of the gases therefrom with the 
oxygen of the barium nitrate; the minimum, the ab- 
sorption of heat expended in decomposition of the 
barium nitrate. 

A comparison of the diagrams shows that the proc- 
esses of combustion in the case of colloids containing 
nitro-glycerin and of those containing metallic ni- 
trates are similar. Both represent aggregates of 
work resulting from successions of independent de- 
compositions. For such powders an element of time 
enters into our conception of chemical action; what 
the ultimate products of combustion are depends 
upon the order of occurrence of successive evolutions 


of various volumes of different gases at high tem- 

The base of the projectile is subjected to a series of 
impulses due to the development of successive waves 
of pressure; the result is an increased initial velocity 
for a given developed pressure, the acceleration be- 
ing sustained throughout a comparatively longer 
period of time. 

Those familiar with experimental development of 
ordnance during recent years remember a type of 
multi-charge gun the construction of which seemed 
based upon a favorable combination of correct prin- 
ciples, but which was rejected on trial, as its practical 
disadvantages were found to outweigh by far its ad- 
vantages. I refer to the Lyman-Haskell multi-charge 
gun, a weapon supplied with a number of pockets dis- 
tributed along the axis of the bore. In each pocket 
a charge of powder was placed; it was supposed that 
the projectile, by uncovering successive pockets in its 
flight, would cause their contents to ignite and thus 
furnish successive accelerating impulses to increase 
its velocity. 

From what has been already said in relation to the 
principle of successive combustion, it will be seen 
that the employment of charges of accelerated pow- 
der, like those above described, in a gun of present- 
day type, represents the limiting extension of the 
multi-charge principle. In relation to their succes- 
sive combustions, the nitro-glycerin and the nitro- 

*See extracts from paper by Prof. Mendel6ef, p. 33. 


cellulose may be considered as sub-charges, con- 
tained in independent chambers or pockets, or dis- 
tributed throughout a very large number of small 

The principle already stated, as established by the 
study of the ballistic action of accelerated or com- 
posite powders, may be now amplified as follows: 

Minute particles of an oxygen-carrier, uniformly in- 
corporated into a nitro-colloid and held in suspension 
throughout the mass of the colloid in an undissolved 
state, act through their independent combustion in such 
a manner as to render more progressive the combustion 
of the colloid into which they are incorporated. 



Accelerator 185 

Acid, nitric, table showing action and results of mixtures of . . . 83 

Nitrohydric 109 

Sulpho-nitrie, table showing action and results of mix- 
tures of 87 

Acid bath, table showing composition of, in relation to yield of 

nitrogen dioxide 132 

Alcohol, absolute, as a solvent 52 

Action of, in facilitating solution 43 

Alkalies, action of, on cellulose 46 

Ammonium compounds as materials for explosives 108 

Ammonium nitrate in smokeless powders 122 

Attainment of maximum propulsive effect 169 

Ballistic efficiencies of pure colloids and compound colloids 

containing accelerators, table showing 167 

Ballistic properties and composition of explosives, table show- 
ing 184, 186 

Ballistite 117 

Benzol derivatives as materials for explosives 121 

Cellulose 5 

Action of alkaline hydrates on 46 

As a base for smokeless powders 125 

Composition of 57 

Conditions governing formulation of 66 

Constitution of 21 

Dinitrate 14 

Electrolytic strain in 50 

Hexanitrate 12 


196 INDEX 


Cellulose hydrate, composition of 57 

Hydration of 45 

Incompletely nitrated 92 

Molecule, theory of 37 

Nitrates *. 74 

Nitrates, composition of 11 

Nitrates, volume of nitrogen dioxide disengaged per gram 

by 128 

Pentanitrate 13 

Structure of 21, 59, 60, 61 

Structure of polymeric forms of 63, 64 

Tetranitrate 13 

Thiocarbonate 48 

Trinitrate 13 

Type 69, 70 

Cocoa powder, gases evolved by 101 

Cold, action of, in accelerating solution 51 

Solubility of nitro-celluloses in the 43 

Solubility of njitro-hydrocellulose in the 41, 42 

Solubility of pyrocellulose in the 43 

Collodion cotton 7 

Collodion-pyroxylin 7, 13 

Collodions 16 

Table showing effect of washing and pulping on viscosity 

of 151 

Colloid 161 

Colloidization 71 

Colloids, accelerated 190 

Colloids, compound, containing accelerators, table showing 

ballistic efficiencies of 167 

Pure, table showing ballistic efficiencies of 167 

Combustion, complete 173 

Of grains 179 

Of powder 110 

Complete combustion 173 

Composite powders 179 

Cordite 166 

Cotton, friable 7 

Nitration of 127 

Researches upon nitration of 81 

Definitions 5 

INDEX 197 


Deterrents 165 

Dinitro- methane, explosive properties of 112 

Ether as a solvent 55 

Ether-alcohol as a solvent 51, 55 

Experiments, various 144 

Explosives, ammonium compounds as materials for 108 

Benzol derivatives as materials for 121 

Hydrocarbons as materials for 109 

Tables showing composition and ballistic properties of. 184, 186 

Friable cottons 7, 16 

Gases, calculating volume evolved 99 

Evolved by ballistite 117 

Evolved by black powder 100 

Evolved by brown powder 101 

Evolved by nitro-cellulose 102 

Evolved by nitroform 113 

Evolved by nitro-glycerin 101, 124 

Evolved by nitrohydric acid 109 

Evolved by pentanitro-cellulose 102 

Evolved by pyrocollodion 105 

Evolved by pyroxylin 122 

Evolved by tetranitro-cellulose 103 

Evolved by trinitro-cellulose 102 

Gun-cotton ..'...' 7, 12, 102 

Discovery of . . 1 

Table showing effect of pulping on 151 

Table showing effect of washing on * 151 

Gun-cottons 16 

Hydration 75 

Hydrocarbons, as materials for explosives 109 


Amorphous 50 

Mercerization 47, 71 

Mononitro-methane, explosive properties of 112 

Multi-charge principle 192 

Nitration 5, 71 

Experiments in 134, 140, 141 

Higher limit of degree of 91 

Of cotton 127 

Of cotton, researches upon 81 

1 98 INDEX 


Nitration, table showing results of, for different periods of 

time 137, 142, 145 

Temperature in 35, 36 

Nitro-cellulose 5, 161 

Composition of 94 

Decomposition-products of 190 

Formula of 94 

Gases evolved by 102 

Insoluble 5 

Of high nitration 5 

Of low nitration 5 

Of mean nitration 5 

Soluble , 6 

Nitro-celluloses, properties of 95 

Solubility of, in the cold 43 

Nitroform, explosive properties of 112 

Gases evolved by 113 

Nitrogen dioxide, table showing yield of, in relation to com- 
position of acid bath 132 

Volume disengaged per gram of cellulose nitrates 128 

Nitre-glycerin 116 

Decomposition-products of 189 

Gases evolved by 101 

Gas volume developed by 124 

In smokeless powders' 182 

Nitro-hydrocellulose 6 

Insoluble 7 

Of high nitration 6 

Of mean nitration 6 

Of low nitration 6 

Soluble 7 

Solubility at freezing temperature 41, 42 

Solutions 38 

Nitro-hydrocelluloses, constitution of 34 

Nitro-mannite 116, 118 

Nomenclature 4 

Oxygen-carriers 32 

Pentanitro-cellulose, gases evolved by 102 

Polymerization 71 

Powder, black, gases evolved by 100 

Cocoa (brown), gases evolved by 101 

INDEX 199 


Powder, pyroxylin 125 

Smokeless, development of 156 

Smokeless, origin of 1 

Smokeless, ammonium nitrate in 122 

Smokeless, cellulose as a base for 125 

Smokeless, nitro-glycerin in 182 

Smokeless, pyrocellulose as a base for 175 

Smokeless, pyrocollodion as a base for 97, 125 

Powders, colloid 28 

Composite 179 

Table showing qualities of 168 

Table showing quantity of nitro- carriers in certain 183 

Propulsive effect, attainment of maximum 169 

Pyrocellulose 7 

Adaptation for smokeless powders 175 

Solubility in the cold 43 

Pyrocollodion 97 

As a base for smokeless powders 125 

Gases evolved by 105 

Smokeless powders from 97 

Pyroxylin 7 

Gas volume developed by 122 

Powder 125 

Pyroxylin and nitro-naphthalin 122 

Pyroxylin and picric acid 122 

Smokeless powder, development of 156 

Origin of 1 

Smokeless powders, ammonium nitrate in 122 

Cellulose as a base for 125 

Nitro-glycerin in 182 

Pyrocellulose as ai base for 175 

Pyrocollodion as a base for. 97, 125 

Solubility, temperature in relation to 41, 44 

Solution, action of cold in accelerating 51 

Table showing action of nitric acid mixtures, and results 83 

Showing action of snlpho-nitrio acid and mixtures, and 

results 87 

Showing ballistic efficiencies of pure colloids and com- 
pound colloids containing accelerators 167 

Showing composition and ballistic properties of explo- 
sives 184, 186 

200 INDEX 


Table showing composition of acid bath in relation to yield 

of nitrogen dioxide 132 

Showing effect of temperature during dipping and reac- 
tion 149 

Showing effect of pulping on gun-cotton 151 

Showing effect of washing on gun-cotton 151 

Showing effect of washing and pulping on the viscosity of 

collodions 151 

Showing qualities of powders 168 

Showing quantity of nitro-carrier in certain powders 183 

Showing results of nitration for different periods of time. 

137, 142, 145 

Temperature, in nitration 35, 36 

In relation to solubility 41, 44 

Table showing effect of, during dipping and reaction 149 

Tetranitro-cellulose, gases evolved by 103 

Tetranitro-methane, explosive properties of 113 

Theory of cellulose molecule 37 

Trinitro- methane, explosive properties of 112 

Viscosity 27 

Of collodions, table showing effect of washing and pulping 
on . . 151 








Descriptive circulars sent on application. 

Books marked with an asterisk are sold at net prices only. 

All books are bound in cloth unless otherwise stated. 



Armsby's Manual of Cattle Feeding. 12mo, $1 75 

Budd and Hausen, American Horticulture Manual .. (In press. ) 

Downing's Fruit and Fruit Trees 8vo, 5 00 

Grotenfelt's The Principles of Modern Dairy Practice. (Woll.) 

12mo, 2 00 

Kemp's Landscape Gardening 12mo, 2 50 

Mayuard's Landscape Gardening 12rno, 1 50 

Steel's Treatise on the Diseases of the Dog 8vo, 3 50 

" Treatise on the Diseases of the Ox 8vo, 6 00 

Stockbridge's Rocks and Soils 8vo, 2 50 

Woll's Handbook for Farmers and Dairymen 12mo, 1 50 



Berg's Buildings and Structures of American Railroads 5 00 

Birkmire's American Theatres Planning and Construction. 8vo, 3 00 

" Architectural Iron and Steel 8vo, 3 50 

" Compound Riveted Girders 8vo, 200 

" Skeleton Construction in Buildings 8vo, 3 00 


Birkmire's Planning and Const ruction of Higli Office Buildings. 

8vo, $3 50 

Briggs' Modern Am. School Building 8vo, 4 00 

Carpenter's Heating and Ventilating of Buildings Svo, 3 00 

Freitag's Architectural Engineering Svo, 2 50 

The Fireproofiug of Steel Buildings 8vo, 2 50 

Gerhard's Sanitary House Inspection 16mo, 1 00 

" Theatre Fires and Panics 12mo, 1 50 

Hatfield's American House Carpenter Svo, 5 00 

Holly's Carpenter and Joiner , .ISmo, 75 

Kidder's Architect and Builder's Pocket-book. . . 16mo, morocco, 4 00 

Merrill's Stones for Building and Decoration Svo, 5 00 

Monckton's Stair Building Wood, Iron, and Stone 4to, 4 00 

Wait's Engineering and Architectural Jurisprudence Svo, 6 00 

Sheep, 6 50 

Worcester's Small Hospitals Establishment and Maintenance, 
including Atkinson's Suggestions for Hospital Archi- 
tecture 12mo, 125 

* World's Columbian Exposition of 1893 Large 4to, 1 00 



*Bruffs Ordnance and Gunnery Svo, 6 00 

Chase's Screw Propellers Svo, 3 00 

Cronkhite's Gunnery for Non-com. Officers 32mo, morocco, 2 00 

* Davis's Treatise on Military Law Svo, 7 00 

Sheep, 7 50 

* " Elements of Law Svo, 250 

De Brack's Cavalry Outpost Duties. (Can-.). . . .32mo, morocco, 2 00 

Dietz's Soldier's First Aid 16mo, morocco, 1 25 

* Dredge's Modern French Artillery Large 4to, half morocco, 15 00 

Record of the Transportation Exhibits Building, 

World's Columbian Exposition of 1893.. 4to, half morocco, 5 00 

Durand's Resistance and Propulsion of Ships Svo, 5 00 

* Fiebeger's Field Fortification, including Military Bridges, 

Demolitions, Encampments and Communications. 

Large 12rno, 2 00 

Dyer's Light Artillery 12mo, 3 00 


Hoff's Naval Tactics 8vo, $1 50 

*Ingalls's Ballistic Tables 8vo, 1 50" 

" Handbook of Problems in Direct Fire 8vo, 4 00 

Mahau's Permanent Fortifications. (Mercur. ). 8 vo, half morocco, 7 50 

* Mercur's Attack of Fortified Places 12mo, 2 00 

* " Elements of the Art t>f War 8vo, 4 00 

Metcalfe's Ordnance and Gunnery 12uio, with Atlas, 5 00 

Murray's A Manual for Courts-Martial IGino, morocco, 1 50 

" Infantry Drill Regulations adapted to the Springfield 

Rifle, Caliber .45 32mo, paper, 10 

* Phelps's Practical Marine Surveying 8vo, 2 50 

Powell's Army Officer's Examiner 12mo, 4 00 

Sharpe's Subsisting Armies 32tno, morocco, 1 50 

Wheeler's Siege Operations 8vo, 2 00 

Wiuthrop's Abridgment of Military Law 12mo, 2 50 

Woodhull's Notes on Military Hygiene 16mo, 1 50 

Young's Simple Elements of Navigation 16mo, morocco, 2 00 

first edition 1 00 



Fletcher's Quant. Assaying with the Blowpipe.. 16mo, morocco, 1 50 

Furman's Practical Assaying 8vo, 3 00 

Kunhardt's Ore Dressing 8vo, 1 50 

Miller's Manual of Assaying 12mo, 1 00 

O'Driscoll's Treatment of Gold Ores 8vo, 2 00 

Ricketts and Miller's Notes on Assaying 8vo, 3 00 

Thurstoii's Alloys, Brasses, and Bronzes 8vo, 2 50 

Wilson's Cyanide Processes .12mo, 1 50 

4 ' The Chlorination Process 12mo, 1 50 



Craig's Azimuth 4to, 3 50 

Doolittle's Practical Astronomy 8vo, 4 00 

Gore's Elements of Geodesy 8vo, 2 50 

Hay ford's Text-book of Geodetic Astronomy .8vo. 3 00 

* Michie and Harlow's Practical Astronomy 8vo, 3 00 

* White's Theoretical and Descriptive Astronomy 12ino, 2 00 




Baldwin's Orchids of New England Small 8vo, $1 50 

Thome's Structural Botany 16mo, 2 25 

Westermaier's General Botany. (Schneider.) 8vo, 2 00 



(See also ENGINEERING, p. 7. ) 

Boiler's Highway Bridges, Svo, 2 00 

* " The Thames River Bridge 4to, paper, 500 

Burr's Stresses in Bridges , Svo, 3 50 

Crehore's Mechanics of the Girder Svo, 5 00 

Dredge's Thames Bridges 7 parts, per part, 1 25 

Du Bois's Stresses in Framed Structures Small 4to, 10 00 

Foster's Wooden Trestle Bridges 4to, 5 00 

Greene's Arches in Wood, etc Svo, 2 50 

Bridge Trusses Svo, 250 

" Roof Trusses Svo, 125 

Howe's Treatise on Arches Svo, 4 00 

Johnson's Modern Framed Structures Small 4to, 10 00 

^[erriman & Jacoby's Text-book of Roofs and Bridges. 

Part I., Stresses Svo, 250 

Merriman & Jacoby's Text-book of Roofs and Bridges. 

Part II., Graphic Statics Svo, 2 50 

Merriman & Jacoby's Text-book of Roofs and Bridges. 

Part III.. Bridge Design Svo, 2 50 

Merriman & Jacoby's Text- book of Roofs and Bridges. 

Part IV., Continuous, Draw, Cantilever, Suspension, and 

Arched Bridges Svo, 2 50 

* Morison's The Memphis Bridge Oblong 4to, 10 00 

Waddell's Iron Highway Bridges , Svo, 4 00 

" De Poutibus (a Pocket-book for Bridge Engineers). 

16mo, morocco, 3 00 

Specifications for Steel Bridges 12mo, 1 25 

Wood's Construction of Bridges and Roofs Svo, 2 00 

Wright's Designing of Draw Spans. Parts I. and II.. Svo, each 2 50 

" " " " Complete Svo, 350 



Adriance's Laboratory Calculations 12mo, $1 25 

Allen's Tables for Iron Analysis 8vo, 3 00 

Austeu's Notes for Chemical Students 12mo, 1 50 

Bolton's Student's Guide in Quantitative Analysis 8vo, 1 50 

Bolt wood's Elementary Electro Chemistry (In the press.) 

Classen's Analysis by Electrolysis. (HerrickandBoltwood.).8vo, 3 00 

Cohu's Indicators and Test-papers 12mo 2 00 

Crafts's Qualitative Analysis. (Schaeffer.) 12mo, 1 50 

Davenport's Statistical Methods with Special Reference to Bio- 
logical Variations 12mo, morocco, 1 25 

Drechsel's Chemical Reactions. (Merrill.) 12mo, 1 25 

Erdmann's Introduction to Chemical Preparations. (Dunlap.) 

12mo, 1 25 

Fresenius's Quantitative Chemical Analysis. (Allen.) 8vo, 6 00 

Qualitative " " (Johnson.) 8vo, 300 

(Wells.) Trans. 

16th German Edition 8vo, 5 00 

Fuertes's Water and Public Health 12mo, 1 50 

Gill's Gas and Fuel Analysis 12mo, 1 25 

Hammarsten's Physiological Chemistry. (Mandel.) 8vo, 4 00 

Helm's Principles of Mathematical Chemistry. (Morgan). 12mo, 1 50 

Hopkins' Oil- Chemist's Hand-book 8vo, 3 00 

Ladd's Quantitative Chemical Analysis 12mo, 1 00 

Landauer's Spectrum Analysis. (Tingle.) 8vo, 3 00 

Lob's Electrolysis and Electrosynthesis of Organic Compounds 1 . 

(Loreuz.) 12mo, 1 00 

Mendel's Bio-chemical Laboratory 12rno, 1 50 

Mason's Water-supply 8vo, 5 00 

' ' Examination of Water 12mo, 1 25 

Meyer's Radicles in Carbon Compounds. (Tingle.) 12mo, 1 00 

Mixter's Elementary Text-book of Chemistry 12mo, 1 50 

Morgan's The Theory of Solutions and its Results 12mo, 1 00 

Elements of Physical Chemistry 12mo, 2 00 

Nichols's Water-supply (Chemical and Sanitary) 8vo, 2 50 

O'Brine's Laboratory Guide to Chemical Analysis 8vo, 2 00 

Pinner's Organic Chemistry. (Austen.) 12mo, 1 50 

Poole's Calorific Power of Fuels 8vo, 3 00 

Richards's Cost of Living as Modified by Sanitary Science.. 12mo. 1 00 

" and Woodman's Air, Water, and Food 8vo, 2 00 

Ricketts and Russell's Notes on Inorganic Chemistry (Non- 
metallic) Oblong 8vo, morocco, 75 

Rideal's Sewage and the Bacterial Purification of Sewage.. ,8vo, 3 50 


Buddiman's Incompatibilities in Prescriptions 8vo, $2 00 

Schimpf's Volumetric Analysis 12mo, 2 50 

Spencer's Sugar Manufacturer's Handbook 16mo, morocco, 2 00 

Handbook for Chemists of Beet Sugar Houses. 

16mo, morocco, 3 00 

Stockbridge's Rocks and Soils 8vo, 2 50 

* Tillman's Descriptive General Chemistry 8vo, 3 00 

Van Deventer's Physical Chemistry for Beginners. (Bolt wood.) 

12mo, 1 50 

"Wells's Inorganic Qualitative Analysis 12mo, 1 50 

" Laboratory Guide in Qualitative Chemical Analysis. 

8vo, 1 50 

Whipple's Microscopy of Drinking-water 8vo, 3 50 

Wiechmann's Chemical Lecture Notes 12mo, 3OO 

" Sugar Analysis Small 8vo, 2 50 

Wulling's Inorganic Phar. and Med. Chemistry 12mo, 2 00 



Hill's Shades and Shadows and Perspective 8vo, 2 00 

MacCord's Descriptive Geometry 8vo, 3 00 

Kinematics 8vo, 500 

" Mechanical Drawing 8vo, 4 00 

Mahan's Industrial Drawing. (Thompson. 1 ) 2 vols., 8vo, 3 50 

Reed's Topographical Drawing. (H. A.) 4to, 5 00 

Reid's A Course in Mechanical Drawing 8vo. 2 00 

" Mechanical Drawing and Elementary Machine Design. 

8vo, 3 00 

Smith's Topographical Drawing. (Macmillan.) 8vo, 250 

Warren's Descriptive Geometry 2 vols., 8vo, 3 50 

Drafting Instruments 12mo, 125 

' ' Free-hand Drawing 12nio, 1 00 

Linear Perspective 12mo, 1 00 

Machine Construction 2vols.,8vo, 750 

Plane Problems 12mo, 125 

" Primary Geometry 12mo, 75 

Problems and Theorems 8vo, 250 

" Projection Drawing 12mo, 150 

" Shades and Shadows ...8vo, 300 

' ' Stereotomy Stone-cutting 8vo, 2 50 

Whelpley's Letter Engraving 12mo, 2 00 

Wilson's Free-hand Perspective 8vo, 2 50 




Anthony and Brackett's Text-book of Physics. (Magie.) Small 

8vo, $3 00 

Anthony's Theory of Electrical Measurements 12mo, 1 00 

Barker's Deep-sea Soundings 8vo, 2 00 

Benjamin's Voltaic Cell 8vo, 3 00 

History of Electricity 8vo, 300 

Classen's Analysis by Electrolysis. (Henick and Bolt\vood j Svo, 3 00 
Crehore and Squier's Experiments with a New Polarizing Photo- 
Chronograph 8vo, 3 00 

Dawson's Electric Railways and Tramways. Small, 4to, half 

morocco, 12 50 

* "Engineering" and Electric Traction Pocket-book. 16mo, 

morocco, 5 00 

* Dredge's Electric Illuminations. . . .2 vols. , 4to, half morocco, 25 00 

Vol. II 4to, 7 50 

Gilbert's De magnate. (Mottelay.) 8vo, 2 50 

Holman's Precision of Measurements 8vo, 2 00 

Telescope-mirror-scale Method Large 8vo, 75 

Lob's Electrolysis and El ectrosyn thesis of Organic Compounds. 

(Lorenz.) 12mo, 1 00 

*Michie's Wave Motion Relating to Sound and Light 8vo, 4 00 

Morgan's The Theory of Solutions and its Results 12mo, 1 00 

Niaudet's Electric Batteries (Fishback.) 12mo, 2 50 

*Parshall & Hobart Electric Generators. Small 4to, half mor., 10 00 

Pratt and Alden's Street-railway Road-beds 8vo, 2 00 

Reagan's Steam and Electric Locomotives 12mo, 2 00 

Thurston's Stationary Steam Engines for Electric Lighting Pur- 
poses 8vo, 250 

*Tillman's Heat 8vo, 1 50 

Tory & Pitcher's Laboratory Physics (In press) 



(See also BRIDGES, p. 4 ; HYDRAULICS, p. 9 ; MATERIALS OF EN- 

Baker's Masonry Construction , .,,,..,.... Svo, 5 00 

Surveying Instruments 12mo, 300 

Black's U. S. Public Works Oblong 4to, 5 00 

Brooks's Street-railway Location 16mo, morocco, 1 50 

Butts's Civil Engineers' Field Book 16mo, morocco, 2 50 

Byrne's Highway Construction Svo, 5 00 


Byrne's Inspection of Materials and Workmanship 16ino, $3 00 

Carpenter's Experimental Engineering 8vo, 6 00 

Church's Mechanics of Engineering Solids and Fluids. .. .Svo, 6 00 

" Notes and Examples in Mechanics Svo, 200 

Crandall's Earthwork Tables 8vo, 1 50 

The Transition Curve 16mo, morocco, 1 50 

* Dredge's Penn. Railroad Construction, etc. Large 4to, 

half morocco, $10; paper, 5 00 

* Drinker's Tunnelling 4to, half morocco, 25 00 

Eissler's Explosives Nitroglycerine and Dynamite Svo, 4 00 

Frizell's Water Power Svo, 5 00 

Folwell's Sewerage Svo, 3 00 

" Water-supply Engineering Svo, 4 00 

Fowler's Coffer-darn Process for Piers . . Svo. 2 50 

Gerhard's Sanitary House Inspection 12nio, 1 00 

Godwin's Railroad Engineer's Field-book 16mo, morocco, 2 50 

Gore's Elements of Geodesy Svo, 2 50 

Howard's Transition Curve Field-book 16mo, morocco, 1 50 

Howe's Retaining Walls (New Edition.) t . . . .12mo, 1 25 

Hudson's Excavation Tables. Vol. II Svo, 1 00 

Hutton's Mechanical Engineering of Power Plants Svo, 5 00 

" Heat and Heat Engines Svo, 5 00 

Johnson's Materials of Construction Large Svo, 6 00 

" Theory and Practice of Surveying Small Svo, 4 00 

Kent's Mechanical Engineer's Pocket-book 16mo, morocco, 5 00 

Kiersted's Sewage Disposal 12mo, 1 25 

Mahan's Civil Engineering. (Wood.) Svo, 5 00 

Merriman and Brook's Handbook for Surveyors. . . .16mo, rnor., 2 00 

Merriman's Precise Surveying and Geodesy Svo, 2 50 

" Sanitary Engineering Svo, 2 00 

Nagle's Manual for Railroad Engineers 16mo, morocco, 3 00 

Ogden's Sewer Design 12mo, 2 00 

Patton's Civil Engineering Svo, half morocco, 7 50 

Patton's Foundations Svo, 5 00 

Philbrick's Field Manual for Engineers 16mo, morocco, 3 00 

Pratt and Aldeu's Street-railway Road-beds Svo, 2 00 

Rockwell's Roads and Pavements in France 12mo, 1 25 

Schuyler's Reservoirs for Irrigation Large Svo. (In press.) 

Searles's Field Engineering . . . . c 16mo, morocco, 3 00 

Railroad Spiral 16mo, morocco, 1 50 

Siebert and Biggin's Modern Stone Cutting and Masonry. . .Svo, 1 50 

Smart's Engineering Laboratory Practice 12rno, 2 50 

Smith's Wire Manufacture and Uses Small 4to, 3 00 

Spalding's Roads and Pavements 12mo, 2 00 

" Hydraulic Cement 12mo, 2 00 


Taylor's Prismoidal Formulas ami Earthwork 8vo, $1 50 

Thurstou's Materials of Construction 8vo, 5 00 

Tillson's Street Pavements and Paving Mateiials 8vo, 4 00 

* Trautwine's Civil Engineer's Pocket-book 16mo, morocco, 5 00 

" Cross-section Sheet, 25 

" Excavations and Embankments 8vo, 2 00 

Laying Out Curves 12mo, morocco, 2 50 

Waddell's De Pontibus (A Pocket-book for Bridge Engineers). 

16mo, morocco, 3 00 

Wait's Engineering and Architectural Jurisprudence 8vo, 6 00 

Sheep, 6 50 

" Law of Field Operation in Engineering, etc 8vo, 5 00 

Sheep, 5 50 

Warren's Stereotomy Stone-cutting. 8vo, 2 50 

Webb's Engineering Instruments. New Edition. 16mo, morocco, 1 25 

Railroad Construction 8vo, 400 

Wegmann's Construction of Masonry Dams 4to, 5 00 

Wellington's Location of Railways. Small 8vo, 5 00 

Wheeler's Civil Engineering 8vo, 4 00 

Wilson's Topographical Surveying 8vo, 3 50 

Wolff's Windmill as a Prime Mover 8vo, 3 00 



(See also ENGINEERING, p. 7.) 
Bazin's Experiments upon the Contraction of the Liquid Vein. 

(Trautwine.) 8vo, 2 00 

Bovey's Treatise on Hydraulics 8vo, 4 00 

Coffin's Graphical Solution of Hydraulic Problems 12mo, 3 50 

Ferrcl's Treatise on the Winds, Cyclones, and Tornadoes. . .8vo, 4 00 

Fol well's Water Supply Engineering 8vo, 4 00 

Fuertes's Water and Public Health 12mo, 1 50 

Ganguillet & Kutter's Flow of Water. (Bering & Trautwine.) 

8vo, 4 00 

Hazen's Filtration of Public Water Supply 8vo, 3 00 

Herschel's 115 Experiments 8vo, 2 00 

Kiersted's Sewage Disposal 12mo, 1 25 

Mason's Water Supply 8vo, 5 00 

" Examination of Water 12mo, 1 25 

Merriman's Treatise on Hydraulics.. 8vo, 4 00 

Nichols's Water Supply (Chemical and Sanitary) 8vo, 2 50 

Turneaure and Russell's Water-supply (In press.) 

Wegmann's Water Supply of the City of New York 4to, 10 00 

Weisbach's Hydraulics. (Du Bois.) 8vo, 5 00 

Whipple's Microscopy of Drinking Water 8vo, 3 50 


Wilson's Irrigation Engineering 8vo, $4 00 

" Hydraulic and Placer Mining 12mo, 2 00 

"Wolff's Windmill as a Prime Mover 8vo, 3 00 

Wood's Theory of Turbines 8vo, 2 50 



Davis's Elements of Law 8vo, 2 50 

4 ' Treatise on Military Law. 8vo, 7 00 

Sheep, 7 50 

Murray's A Manual for Courts-martial IGmo, morocco, 1 50 

Wait's Engineering and Architectural Jurisprudence 8vo, 6 00 

Sheep, 6 50 

" Laws of Field Operation in Engineering 8vo, 5 00 

Sheep, 5 50 

Winthrop's Abridgment of Military Law 12nio, 2 50 



Allen's Tables for Iron Analysis 8vo, 3 00 

Beaumont's Woollen and Worsted Manufacture 12ino, 1 50 

Bolland's Encyclopaedia of Founding Terms 12ino, 3 00 

The Iron Founder 12mo, 250 

Supplement 12mo, 250 

Bouvier's Handbook on Oil Painting 12mo, 2 00 

Eissler's Explosives, Nitroglycerine, and Dynamite 8vo, 4 00 

Ford's Boiler Making for Boiler Makers 18mo, 1 00 

Metcalfe's Cost of Manufactures 8vo, 5 00 

Melcalf 's Steel A Manual for Steel Users 12mo, 2 00 

*Reisig's Guide to Piece Dyeing 8vo, 25 00 

Spencer's Sugar Manufacturer's Handbook IGmo, morocco, 2 00 

" Handbook for Chemists of Beet Sugar Houses. 

16mo, morocco, 3 00 

Thurston's Manual of Steam Boilers 8vo, 5 00 

Walke's Lectures on Explosives 8vo, 4 00 

West's American Foundry Practice 12mo, 2 50 

" Moulder's Text-book 12mo, 2 50 

Wiechmaun's Sugar Analysis Small 8vo, 2 50 

Woodbury's Fire Protection of Mills 8vo, 2 50 


(See also ENGINEERING, p. 7.) 

Baker's Masonry Construction 8vo, 5 00 

Beardslee and Kent's Strength of Wrought Iron. 8vo, 1 50 


Bovey's Strength of Hateriuls * 8vo, 7 50 

Burr's Elasticity and Resistance of Materials 8vo, $5 00 

BjTne's Highway Construction 8vo, 5 00 

Church's Mechanics of Engineering Solids and Fluids 8vo, 6 00 

Du Bois's Stresses in Framed Structures Small 4to, 10 00 

Johnson's Materials of Construction 8vo, 6 00 

Lanza's Applied Mechanics. 8vo, 7 50 

Martens's Testing Materials. (Henning.) 2 vols., 8vo, 7 50 

Merrill's Stones for Building and Decoration Svo, 5 00 

Merriinan's Mechanics of Materials 8vo, 4 00 

" Strength of Materials 12rno, 1 00 

Pattou's Treatise on Foundations 8vo, 5 00 

Rockwell's Roads and Pavements in France 12mo, 1 25 

Spalding's Roads and Pavements 12mo, 2 00 

Thurston's Materials of Construction 8vo, 5 00 

Materials of Engineering . 3 vols., 8vo, 8 00 

Vol. I. , Non-metallic 8vo, 2 00 

Vol. II., Iron and Steel 8vo, 3 50 

Vol. Ill:, Alloys, Brasses, and Bronzes 8vo, 2 50 

Wood's Resistance of Materials 8vo, 2 00 



Baker's Elliptic Functions 8vo, 1 50 

*Bass's Differential Calculus 12mo, 4 00 

Briggs's Plane Analytical Geometry 12mo, 1 00 

Chapman's Theory of Equations 12mo, 1 50 

Compton's Logarithmic Computations 12mo, 1 50 

Davis's Introduction to the Logic of Algebra 8vo, 1 50 

Halsted's Elements of Geometry .Svo, 1 75 

" Synthetic Geometry 8vo, 1 50 

Johnson's Curve Tracing 12mo, 1 00 

" Differential Equations Ordinary and Partial. 

Small Svo, 3 50 

" Integral Calculus 12mo, 1 50 

" " " Unabridged. Small Svo. (In pi-ess.) 

" Least Squares 12mo, 1 50 

*Ludlow's Logarithmic and Other Tables. (Bass.) Svo, 2 00 

* ' ' Trigonometry with Tables. (Bass.) Svo, 3 00 

*Mahan's Descriptive Geometry (Stone Cutting) Svo, 1 50 

Merriman and Woodward's Higher Mathematics Svo, 5 00 

Merrirnan's Method of Least Squares Svo, 2 00 

Rice and Johnson's Differential and Integral Calculus, 

2 vols. in 1, small Svo, 2 50 

Rice L~d Johnson's Differential Calculus Small 8vo, $3 00 

Abridgment of Differential Calculus. 

Small 8vo, 1 50 

Totten's Metrology 8vo, 2 50 

Warren's Descriptive Geometry 2 vols., 8vo, 3 50 

' ' Drafting Instruments 12mo, 1 25 

" Free-band Drawing 12mo, 100 

Linear Perspective 12mo, 1 00 

" Primary Geometry 12mo, 75 

Plane Problems 12mo, 1 25 

" Problems and Theorems 8vo, 2 50 

Projection Drawing 12mo, 1 50 

Wood's Co-ordinate Geometry. ....... , . .8vo, 2 00 

Trigonometry 12mo, 100 

Woolf's Descriptive Geometry Large 8vo, 3 00 


(See also ENGINEERING, p. 8.) 

Baldwin's Steam Heating for Buildings 12mo, 2 50 

Barr's Kinematics of Machinery 8vo, 2 50 

Benjamin's Wrinkles and Recipes 12mo, 2 00 

Chordal's Letters to Mechanics 12mo, 2 00 

Church's Mechanics of Engineering 8vo, 6 00 

" Notes and Examples in Mechanics 8vo, 2 00 

Crehore's Mechanics of the Girder 8vo, 5 00 

Cromwell's Belts and Pulleys 12mo, 1 50 

Toothed Gearing 12mo, 1 50 

Comptou's First Lessons in Metal Working 12mo, 1 50 

Compton and De Groodt's Speed Lathe 12mo, 1 50 

Dana's Elementary Mechanics 12mo, 1 50 

Dingey's Machinery Pattern Making 12mo, 2 00 

* Dredge's Trans. Exhibits Building, World Exposition. 

Large 4to, half morocco, 5 00 

Du Bois's Mechanics. Vol. I., Kinematics 8vo, 3 50 

Vol. II., Statics 8vo, 400 

Vol. III., Kinetics 8vo, 350 

Fitzgerald's Boston Machinist 18mo, 1 00 

Flather's Dynamometers 12mo, 2 00 

Rope Driving 12mo, 200 

Hall's Car Lubrication 12mo, 1 00 

Holly's Saw Filing 18mo, 75 

Johnson's Theoretical Mechanics. An Elementary Treatise. 
(In the press.} 

Jones's Machine Design. Part L, Kinematics 8vo, 1 50 


Jones's Machine Design. Part II., Strength and Proportion of 

Machine Parts 8vo, $3 00 

Lanza's Applied Mechanics 8vo, 7 50 

MacCord's Kinematics 8vo, 5 00 

Merriman's Mechanics of Materials 8vo, 4 00 

Metcalfe's Cost of Manufactures 8vo, 5 00 

*Michie's Analytical Mechanics 8vo, 4 00 

Richards's Compressed Air 12mo, 1 50 

Robinson's Principles of Mechanism 8vo, 3 00 

Smith's Press-working of Metals 8vo, ft 00 

Thurston's Friction and Lost Work 8vo, 3 00 

" The Animal as a Machine 12mo, 100 

Warren's Machine Construction 2 vols., 8vo, 7 50 

Weisbach's Hydraulics and Hydraulic Motors. (Du Bois.)..8vo, 5 00 
" Mechanics of Engineering. Vol. III., Part L, 

Sec. I. (Klein.) , 8vo, 500 

Weisbach's Mechanics of Engineering. Vol. III., Part I., 

Sec. II. (Klein.) 8vo, 500 

Weisbach's Steam Engines. (Du Bois.) 8vo, 500 

Wood's Analytical Mechanics 8vo, 3 00 

Elementary Mechanics 12mo, 125 

" " " Supplement and Key 12mo, 1 25 



Allen's Tables for Iron Analysis 8vo, 3 00 

Egleston's Gold and Mercury Large 8vo, 7 50 

" Metallurgy of Silver ' Large 8vo, 750 

* Kerl's Metallurgy Copper and Iron 8vo, 15 00 

* " " Steel, Fuel, etc 8vo, 1500 

Kunhardt's Ore Dressing in Europe 8vo, 1 50 

Metcalf 's Steel A Manual for Steel Users 12mo, 2 00 

O'Driscoll's Treatment of Gold Ores 8vo, 2 00 

Thurston's Iron and Steel 8vo, 3 50 

Alloys 8vo, 250 

Wilson's Cyanide Processes 12mo, 1 50 



Barringer's Minerals of Commercial Value Oblong morocco, 2 50 

Beard's Ventilation of Mines 12mo, 2 50 

Boyd's Resources of South Western Virginia 8vo, 3 00 

" Map of South Western Virginia Pocket-book form, 2 00 

Brush and Penfield's Determinative Mineralogy. New Ed. 8vo, 4 00 


Chester's Catalogue of Minerals 8vo, $1 25 

" " " " Paper, 50 

" Dictionary of the Names of Minerals 8vo, 3 00 

Dana's American Localities of Minerals. Large 8vo, 1 00 

" Descriptive Mineralogy (E. S.) Large 8vo. half morocco, 1250 

" First Appendix to System of Mineralogy. . . .Large 8vo, 1 00 

Mineralogy and Petrography. (J. D.) 12mo, 2 00 

" Minerals and How to Study Them. (E. S.) 12mo, 1 50 

" Text-book of Mineralogy. (E. S.).. .New Edition. 8vo, 400 

* Drinker's Tunnelling, Explosives, Compounds, and Rock Drills. 

4to, half morocco, 25 00 

Egleston's Catalogue of Minerals and Synonyms 8vo, 2 50 

Eissler's Explosives Nitroglycerine and Dynamite 8vo, 4 00 

Hussak's Rock- forming Minerals. (Smith.) Small 8vo, 2 00 

Ihlseng's Manual of Mining 8vo, 4 00 

Kunhardt's Ore Dressing in Europe , 8vo, 1 50 

O'Driscoll's Treatment of Gold Ores 8vo, 2 00 

* Penfield's Record of Mineral Tests Paper, 8vo, 50 

Roseubusch's Microscopical Physiography of Minerals and 

Rocks. (Iddiugs.) 8vo, 500 

Sawyer's Accidents in Mines Large 8vo, 7 00 

Stockbridge's Rocks and Soils 8vo, 2 50 

*Tillman's Important Minerals and Rocks Svo, 2 00 

Walke's Lectures on Explosives Svo, 4 00 

Williams's Lithology Svo, 3 00 

Wilson's Mine Ventilation 12rno, 1 25 

" Hydraulic and Placer Mining 12ino, 2 50 


(See also ENGINEERING, p. 7.) 

Baldwin's Steam Heating for Buildings 12mo, 2 50 

Clerk's Gas Engine , Small Svo, 400 

Ford's Boiler Making for Boiler Makers 18mo, 1 00 

Hemen way's Indicator Practice 12mo, 2 00 

Kneass's Practice and Theory of the Injector Svo, 1 50 

MacCord's Slide Valve Svo, 2 00 

Meyer's Modern Locomotive Construction 4to, 10 00 

Peabody and Miller's Steam-boilers Svo, 4 00 

Peabody's Tables of Saturated Steam Svo, 1 00 

" Thermodynamics of the Steam Engine Svo, 5 00 

" Valve Gears for the Steam Engine Svo, 2 50 

" Manual of the Steam-engine Indicator . .12rno, 1 50 

Fray's Twenty Years with the Indicator Large Svo, 2 50 

Pupin and Osterberg's Thermodynamics , . .12mo, 1 25 


Reagan's Steam and Electric Locomotives 12mo, $2 00 

Routgeu's Thermodynamics. (Du Bois. ) 8vo, 5 00 

Sinclair's Locomotive Running 12mo, 2 00 

Snow's Steam-boiler Practice 8vo. 3 00 

Thurston's Boiler Explosions 12mo, 1 50 

Engine and Boiler Trials 8vo, 500 

" Manual of the Steam Engine. Part I., Structure 

and Theory, 8vo, 6 00 

" Manual of the Steam Eugine. Part II., Design, 

Construction, and Operation 8vo, 6 00 

2 parts, 10 00 

Thurston's Philosophy of the Steam Engine 12mo, 75 

Reflection on the Motive Power of Heat. (Caruot.) 

12mo, 1 50 

Stationary Steam Engines 8vo, 2 50 

" Steam-boiler Construction and Operation 8vo, 5 00 

Spaugler's Valve Gears 8vo, 2 50 

Weisbach's Steam Engine. (Du Bois.) 8vo, 5 00 

Whitham's Steam-engine Design 8vo, 5 00 

Wilson's Steam Boilers. (Flather.) 12mo, 2 50 

Wood's Thermodynamics, Heat Motors, etc 8vo, 4 00 



Adriance's Laboratory Calculations 12mo, 1 25 

Allen's Tables for Iron Analysis 8vo, 3 00 

Bixby's Graphical Computing Tables Sheet, 25 

Comptou's Logarithms 12mo, 1 50 

Crandall's Railway and Earthwork Tables 8vo, 1 50 

Egleston's Weights and Measures 18mo, 75 

Fisher's Table of Cubic Yards Cardboard, 25 

Hudson's Excavation Tables. Vol. II 8vo, 1 00 

Johnson's Stadia and Earthwork Tables , 8vo, 1 25 

Ludlow's Logarithmic and Other Tables. (Bass.) 12mo, 2 00 

Totteu's Metrology - 8vo, 2 50 



Baldwin's Steam Heating 12mo, 2 50 

Beard's Ventilation of Mines 12mo, 2 50 

Carpenter's Heating and Ventilating of Buildings 8vo, 3 00 

Gerhard's Sanitary House Inspection 12mo, 1 00 

Wilson's Mine Ventilation 12mo, I 25 



Alcott's Gems, Sentiment, Language Gilt edges, $5 00 

Davis's Elements of Law 8vo, 2 00 

Emmon's Geological Guide-book of the Rocky Mountains. .8vo, 1 50 

Ferrel's Treatise on the Winds 8vo, 4 00 

Haines's Addresses Delivered before the Am. Ry. Assn. ..12mo, 2 50 

Mott's The Fallacy of the Present Theory of Sound. .Sq. IGmo, 1 00 

Richards's Cost of Living 12mo, 1 00 

Ricketts's History of Rensselaer Polytechnic Institute 8vo, 3 00 

Rotherham's The New Testament Critically Emphasized. 

12mo, 1 50 
" The Emphasized New Test. A new translation. 

Large 8vo, 2 00 

Totten's An Important Question in Metrology 8vo, 2 50 

* Wiley's Yosemite, Alaska, and Yellowstone 4to, 3 00 



Gesenius's Hebrew and Chaldee Lexicon to Old Testament. 

(Tregelles. ) Small 4to, half morocco, 5 00 

Green's Elementary Hebrew Grammar 12mo, 1 25 

Grammar of the Hebrew Language (New Edition ).8vo, 3 00 

" Hebrew Chrestomathy 8vo, 2 00 

Letteris's Hebrew Bible (Massoretic Notes in English). 

8vo, arabesque, 2 25 


Hammarsten's Physiological Chemistry. (Mandel.) 8vo, 4 00 

Mott's Composition, Digestibility, and Nutritive Value of Food. 

Large mounted chart, 1 25 

Ruddiman's Incompatibilities in Prescriptions 8vo, 2 00 

Steel's Treatise on the Diseases of the Ox 8vo, 6 00 

" Treatise on the Diseases of the Dog 8vo, 3 50 

Woodhull's Military Hygiene 16mo, 1 50 

Worcester's Small Hospitals Establishment and Maintenance, 
including Atkinson's Suggestions for Hospital Archi- 
tecture 12mo, 1 25 



Return to desk from which borrowed. 
This book is DUE on the last date stamped below. 

FEB 4 19 48 





^*-,~'n l-O 

JAN IU 1959 


DEC1 * 

SEP 0.275 

LD 21-100m-9,'47(A5702sl6)476